Artificial neural network and mathematical modeling for Congo red dye remediation using acid-activated mixed waste biomass

The study explores the potential of novel acid-activated algal – pineapple peel biomass (AAPPB) for the removal of Congo red dye with artificial intelligence-based predictive modeling. The characterization analysis confirmed the better surface and functional nature of AAPPB. Batch parameter studies revealed an optimal dose of 1 g/L with a contact time of 40 min. Isotherm and kinetic modeling analysis inferred Redlich-Peterson and Pseudo-second order model to be the best fit, indicating the monolayer, heterogeneous, and chemisorption nature. Maximal adsorption removal ability of 152.3 mg/g was observed from isotherm analysis for AAPPB. Thermodynamic analysis inferred the interaction between AAPPB and Congo red dye molecules to be spontaneous, favourable, and exothermic. A predictive model using Artificial Neural Network (ANN) achieved a correlation coefficient of 0.9943 with ANN testing demonstrating strong agreement with experimental results, confirming the ANN model's reliability in estimating Congo red dye removal by AAPPB. The study provides a sustainable adsorbent with strong potential for dye remediation applications.