Experimental and artificial neural network modelling of Rhodamine B decolorization at low concentrations using pine needle and sugarcane bagasse biochars

The present experimental research investigation accentuates the thermochemical conversion of pine needle, and sugarcane bagasse biomass feedstocks into biochars via facile pyrolysis for the remediation of Rhodamine B (RhB) dye contaminated effluent at lower concentrations accompanied by Artificial Neural Network (ANN) Modelling. The physicochemical properties of the as-synthesized biochars were analyzed using advanced characterization techniques. At adsorbate concentrations of 1 mg L⁻¹, pine needle biochar (PNB) demonstrated a higher maximum adsorption capacity (q_m) of 0.8370 mg/g in contrast to sugarcane bagasse biochar (SBB - q_m: 0.5657 mg/g), indicating enhanced adsorptive site availability and surface functionality. RhB decolorization phenomenon using both the biomass-derived biochars could be best explicated by the Pseudo second-order kinetics, and Langmuir isotherm model, signifying chemisorption, and mono-layered adsorption respectively. The dye adsorption process modelled using the Levenberg Marquardt backpropagation Algorithm optimized via Genetic Algorithm demonstrated a strong correlation with the experimentally obtained RhB adsorption capacity values. Higher coefficient of determination (R²- PNB: 0.987, SBB: 0.9953), and correlation coefficient (R– PNB: 0.9764, SBB: 0.9899) values, along with lower mean square error (MSE- PNB: 0.0050, SBB: 0.0017) values were obtained towards RhB uptake using the biochar adsorbents, signifying favorable predictiveness towards multi-factor adsorption. The optimal parameters for RhB uptake using both the biomass-derived biochars were ascertained to be: temperature (298.15 K), biochar dose (0.02 g), contact time (360 min), pH (2), preliminary RhB concentration (1 mg L⁻¹).