Efficient eco-friendly porous green carbon fiber derived from palm waste leaves for lead removal from wastewater: RSM, and advanced modeling of machine learning

Abstract

This study introduces a novel adsorbent material green carbon fibers (GCF) synthesized from palm waste leaves for sustainable wastewater treatment, leveraging their high porosity, thermal stability, and surface functionality for effective lead (Pb²⁺) removal from aqueous media. The uniqueness of the material lies in its origin from agricultural waste and its retention of natural fibrous morphology and hierarchical pore structures, distinguishing it from conventional activated carbons. Experimental conditions included pH 5.5, temperatures (10–40) °C, adsorbent doses (25–75) mg, and lead ion concentrations (50–100) mg/L. Characterization using SEM, EDS, XRD, TGA, BET, FTIR, and zeta potential analyses confirmed a highly porous GCF structure and favorable surface chemistry. Adsorption kinetics followed a pseudo-second-order (PSO) model (R² = 0.952), indicating chemisorption, with a maximum capacity of 74.961  mg/g. Freundlich isotherm modeling (R² = 0.9646) suggested heterogeneous multilayer adsorption. The RSM model demonstrated high predictive reliability with Reduced 2FI model (R² = 0.9652) and among machine learning models, the Decision Tree outperformed others with R² = 0.9143 and MSE = 76.88, identifying adsorbent weight as the most critical factor. These findings highlight the viability of converting agricultural waste into high-performance adsorbents, offering an eco-friendly and cost-effective solution for heavy metal remediation in wastewater.