A crossbred mesoporous composite of Ni0.15-doped Zn0.85S @ alginate-derived graphitic carbon beads for methylene blue removal: DFT and machine learning investigations

In this study, we developed a novel mesoporous composite material composed of nickel-doped zinc sulfide (Nidoped ZnS) nanoplatelets embedded within alginate-derived graphitic carbon beads (MAGCBs) for the efficient
removal of methylene blue (MB) from aqueous solutions. The composite was characterized using SEM, TEM,
XRD, XPS, BET, and zeta potential analyses, revealing a well-structured mesoporous framework with uniformly
distributed Ni-doped ZnS nanoplatelets, enhancing the surface area and providing abundant active sites for
adsorption. The adsorption performance of MAGCBs reached a maximum removal efficiency of 99.5 % at neutral
pH, significantly outperforming unmodified beads (78 %). Density Functional Theory (DFT) calculations
confirmed that Ni-doping enhanced electrostatic attractions, π–π interactions, and coordination bonding between
MB and the MAGCB surface. The adsorption energies confirmed the thermodynamic stability of MB adsorption
on MAGCB (− 94.79 kcal/mol) compared to AGCB (− 87.08 kcal/mol). Machine learning models were also
applied to predict adsorption efficiency based on key parameters such as pH, stirring time, adsorbent dose, and
initial MB concentration. The Extra Trees algorithm provided the most accurate predictions, identifying pH and
stirring time as the most influential factors on adsorption efficiency. Furthermore, MAGCBs demonstrated
excellent reusability, maintaining over 91 % of their adsorption capacity after ten cycles, making them a highly
promising candidate for large-scale wastewater treatment applications. These findings highlight the effective­
ness, cost-efficiency, and reusability of MAGCBs for dye removal from polluted water, with valuable insights into
the adsorption mechanisms provided by DFT and machine learning models.