Experimental and Machine Learning Modeling of Ni(II) Ion Adsorption onto Guar Gum: ANN and KNN Comparative Study

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

This study employed Guar Gum (GG) to decontaminate synthetic wastewater containing Ni(II) ions. The adsorbent was characterized using FTIR, scanning electron microscopy (SEM), thermogravimetric analysis (TGA), BET surface area, and X-ray diffraction (XRD). The effects of key operating parameters, contact time, pH, sorbent mass, temperature, and initial Ni(II) concentration, were systematically investigated. The optimum conditions for Ni(II) removal were found to be 0.8 g/50 mL adsorbent dosage, pH 7.0, and a contact time of 30 minutes. In addition to single-factor studies, a hierarchical cluster analysis (dendrogram) was performed to evaluate the relative influence of the process variables on removal efficiency. The analysis revealed that adsorbent dosage and pH were the most dominant factors, clustering closely with removal efficiency, while contact time and temperature formed a secondary group, and initial concentration appeared as a distinct parameter. This multivariate insight highlights the central role of pH and dosage in optimizing Ni(II) adsorption. To further predict adsorption performance, the experimental data were employed to develop an Artificial Neural Network (ANN) model. The model achieved a correlation coefficient (R²) of 0.967 and a mean square error (MSE) of 3.857, confirming strong predictive capability. In addition, K-Nearest Neighbors (KNN) modeling was applied and compared with ANN, further validating the predictive performance of the adsorption process. The kinetic analysis demonstrated that the pseudo-second-order model provided the best description of the adsorption kinetics. At equilibrium, the data were consistent with the Langmuir isotherm, yielding a maximum adsorption capacity of 86.0 mg g⁻¹. Thermodynamic analysis further indicated that the process is predominantly governed by physisorption and proceeds exothermically.

Article activity feed