Tannery shaving dust-based charcoal blended adsorbent for efficient heavy metal remediation: An experimental and machine learning approach

Abstract

Tannery shaving dust (TSD) is one of the tannery wastes, poses significant concerns due to its availability and carcinogenic properties. This study has focused on utilizing this waste as adsorbent for heavy metals (HMs) treatment in wastewater. By crosslinking polyvinyl alcohol (PVA) with activated tannery shaving dust (AsD) from TSD and modified charcoal (MC) powder, a composite PVA-AsD-MC adsorbent (PAsMc) was fabricated to remove As, Cr, Zn and Pb from synthetic wastewater. Here, PVA-AsD (1:10) blended with 2:3 MC has sufficient active sites that were ensured by the FT-IR. As an adsorbent the PAsMc showed more thermal stability than AsD, and surface morphology was observed as highly rough. Moreover, the batch experiments have considered pH, adsorbent dose, and contact time factors, achieving impressive metals removal efficiencies: 98.86 % for As, 99.45 % for Cr, 99.72 % for Zn, and 98.30 % for Pb. The optimal conditions were identified as an adsorbent dosage of 4.0 g/L for 25 minutes, and an agitation speed of 300 rpm at pH 8.0–9.0. The adsorption isotherm and kinetics model provided an auspicious result for chemisorption adsorption on the surface. Notably, these datasets were then enhanced with the machine learning model, specifically the Random Forest (RF), aimed at predicting the removal of HMs. The R² values for the training and testing dataset within the range of 0.9927–0.9984 and 0.9940–0.9975 with a RMSE value of 0.9622–1.4612 and 1.1125–1.9294, respectively. Ultimately, a predictive model for HMs removal was developed, which will assist in making rational applications of PAsMc in wastewater treatment.