Integrating Artificial Neural Networks and Response Surface Methodology for Predictive Modeling and Mechanistic Insights into the Detoxification of Hazardous MB and CV Dyes Using Saccharum Officinarum L. Biomass.

The presence of dye pollutants in industrial wastewater poses significant environmental and health risks, necessitating effective treatment methods. The optimal adsorption treatment of methylene blue (MB) and crystal violet (CV) dye-simulated wastewater utilising Saccharum officinarum L presents a key challenge in the selection of appropriate modelling approaches. While RSM and ANN models are frequently used, there is a noticeable knowledge gap when it comes to evaluating their relative strengths and weaknesses in this context. The study compared the predictive abilities of response surface methodology (RSM) and artificial neural network (ANN) for the adsorption treatment of MB and CV dye-simulated wastewater using Saccharum officinarum L. The process experimental variables were modelled and predicted using a three-layer artificial neural network trained using the Levenberg-Marquard backpropagation algorithm and 30 central composite designs (CCD). The adsorption study used a specific mechanism, which led to noteworthy maximum removals of 98.3% and 98.2% for dyes (MB and CV), respectively. The RSM model achieved an impressive R2 of 0.9417, while the ANN model achieved 0.9236 in MB. Adsorption is commonly used to remove colour from many different materials. Saccharum officinarum L., a byproduct of sugarcane processing, has shown potential as an efficient and ecological adsorbent in this environment. The purpose of this study is to evaluate sugarcane bagasse's potential as an adsorbent for the removal of dyes MB and CV from industrial wastewater, providing a long-term strategy for reducing dye pollution. Due to its beneficial economic and environmental characteristics, the Saccharum officinarum L. adsorbent has prompted research into sustainable resources with low pollutant indices.