SAP Nanomicelle-Functionalized Biochar: A Multifunctional and Sustainable Adsorbent for Efficient Removal of Dyes and Heavy Metals.

Herein, a novel multifunctional eco-adsorbent (SMBC) was developed by directly anchoring self-assembled Sapindus saponin (SAP) nanomicelles onto biochar (BC) through a simple and effective method. The resulting material exhibited a significantly enhanced adsorption capacity for p-nitrophenol (PNP), showing a 48.83% improvement compared to unmodified BC. The exceptional adsorption performance for PNP is attributed mainly to polar interactions, π-π interactions, and hydrogen bonding, resulting in a maximum adsorption capacity of 111.21 mg/g. Moreover, SMBC demonstrated high adsorption capacities for cadmium (Cd-(II)), lead (Pb-(II)), methylene blue (MB), and bisphenol A (BPA), reaching 100.17, 104.59, 108.64, and 98.90 mg/g, respectively, surpassing most previously reported adsorbents. Notably, SMBC maintained robust removal performance even in binary systems containing PNP. Characterization analyses indicated that SMBC possesses a greater number of active sites, a more amorphous structure, and a higher abundance of functional groups compared to BC. The adsorption of PNP onto SMBC was determined to be spontaneous, favorable, and exothermic. Equilibrium isotherms were best fitted by the Sips model, while kinetics were well described by a pseudo-second-order model. The adsorption process was minimally affected by ionic strength, coexisting impurities, or changes in pH, and SMBC retained over 80% of its adsorption efficiency after five regeneration cycles. In addition, the dosage of SMBC required to achieve 99% removal of PNP was predicted across a range of solution volumes. Six machine learning models were applied to predict the adsorption of PNP onto SMBC, with Shapley Additive exPlanations (SHAP) utilized for interpretability. Among these, Extreme Gradient Boosting (XGBoost) delivered the best predictive performance, and feature importance was ranked in the following order: Time > Ratio > Concentration > Temperature > Dosage > pH. Cost analysis confirmed that SMBC is a highly attractive and cost-effective eco-adsorbent, suitable for removing diverse pollutants. The accurate machine learning model offers valuable insight into the design of adsorbents and the optimization of operational parameters in practical applications.