Sustainable methylene blue dye removal via bio-derived micro/micron-sized porous particles Zygophyllum coccineum and Calotropis procera: A machine learning-assisted study.

The discharge of synthetic dyes into aquatic ecosystems poses a critical threat to environmental and human health, necessitating urgent, cost-effective, and sustainable remediation strategies. This study addresses a practical yet underexplored question in biosorption research: Is the additional energy and processing cost of converting plant-based materials from micro- to Micron-sized porous particles scale justified by a significantly higher dye removal performance? To answer this, we systematically investigate two underutilized and readily available plant species—Zygophyllum coccineum (Zygo) and Calotropis procera (Cal)—as biosorbents for methylene blue (MB) removal from wastewater. Biosorbents at both micro- and Micron-sized porous particles scale were prepared via mechanical ball milling and thoroughly characterized using SEM, XRD, BET, FTIR, and TGA to ensure powerful structural and compositional validation. Ball milling is overall an environmentally friendly and resource-efficient preparation method, facilitating solid waste recycling and cleaner synthesis with lower chemical emissions. The main environmental and economic drawback is relatively high energy consumption, particularly for large-scale or prolonged operations, but this can be managed through process optimization and proper equipment maintenance. When integrated with energy-saving strategies and dust control, ball milling provides a sustainable, circular alternative to conventional, more polluting approaches to material processing and biosorbent preparation. Characterization confirmed enhanced surface area and porosity in Micron-sized porous particles scale materials (e.g., Micron-sized porous particles-Cal: 86.91 m²/g), directly correlating with increased adsorption efficiency. Adsorption experiments, optimized across batch experiments (dose: 0.5 g/L, pH 7, time: 180 min), demonstrated that Micron-sized porous particles-Cal achieved up to 99.5% MB removal at room temperature, significantly outperforming its microscale equivalent and many reported biosorbents. Adsorption data fitting to Langmuir and Freundlich isotherms, and pseudo-second-order kinetics confirmed the presence of both monolayer adsorption and chemisorption-driven mechanisms. Importantly, we integrated an XGBoost machine learning model to predict adsorption performance with high accuracy (MSE: 0.08, RMSE: 0.28, MAE: 0.15), revealing dosage and pH as dominant factors. This predictive modeling framework not only validated experimental findings but also demonstrated its potential for minimizing trial-and-error in process optimization. By highlighting comparative between processing cost and performance gain, and embedding predictive analytics into material design, this work offers a rigorous, scalable, and environmentally sustainable blueprint for dye removal. It establishes a decision-making basis for whether Micron-sized porous particles-sizing of plant biosorbents is scientifically and industrially justified, contributing both scientific insight and practical guidance to the field of wastewater treatment.