Machine learning and modelling approach for removing methylene blue from aqueous solutions: Optimization, kinetics and thermodynamics studies

Abstract

Background

The present study employs the Juglans Regia, a natural adsorbent, to investigate its efficiency in methylene blue (MB) removal from aqueous solutions. Advanced techniques like Artificial Neural Networks (ANN) and Response Surface Methodologies (RSM) are applied to model and predict the adsorptive behaviour of MB using Juglans Regia adsorbent. Different characterization techniques are utilised to understand the morphology and structure of the catalyst to provide insights into its potential adsorption capabilities.

Methods

The study uses ANN and RSM for predictive modelling, enabling a comprehensive comparison of their effectiveness in describing MB adsorption onto Juglans Regia. Kinetic analysis employing pseudo-2nd order kinetics and isotherm studies based on the Langmuir isotherm model are conducted to understand the kinetics and equilibrium behaviour of the adsorption process. The thermodynamics of the adsorption process are investigated to ascertain the spontaneity and exothermic nature of MB dye removal by Juglans Regia.

Significant Findings

Both ANN and RSM models are predicted the high adsorption efficiency, reaching up to 94.6 and 93.2 %, demonstrating the effectiveness of Juglans Regia as an adsorbent for MB removal. RSM and ANN models are strongly associated with R² values of 0.9117 and 0.9373, respectively. Low Root mean square error (RMSE) and Hybrid fractional error function (HYBRID) computed error function values revealed good agreement between experimental results and model predictions. The study successfully validates the optimization process, leading to the maximum predicted adsorption efficiency values, and providing insights into optimal conditions for efficient MB removal. Thermodynamic analysis confirms that the adsorption process is exothermic and spontaneous, further supporting the potential of Juglans Regia in wastewater remediation of organic pollutants.