Methylene blue

This study investigates the adsorption of methylene blue (MB) from aqueous solution using citrus peel biochar (CPB) and develops a predictive modeling framework based on Artificial Neural Networks (ANN) for process optimization. CPB was prepared by carbonization process under limited oxygen at 750 °C and 10 °C/min heating rate. A yield of 21.30 ± 0.5% was obtained. The prepared biochar is characterized by several techniques. FTIR analysis demonstrates the presence of the characterized functional groups on the biochar surface. The thermogravimetric analysis of biochar shows three main stages of weight loss. CPB exhibits relatively high thermal stability due to the formation of a carbon-rich structure after carbonization. SEM images reveal that the produced biochar exhibits a highly heterogeneous and porous surface morphology characterized by an irregular sponge-like architecture with numerous cavities and interconnected channels. The specific surface area of CPB was indirectly estimated using the methylene blue adsorption method, yielding a value of 242.81m²/g; this value represents an approximation of the accessible aqueous surface area. Batch adsorption experiments were conducted to evaluate the effects of pH, contact time, temperature, adsorbent dosage, and initial dye concentration on removal efficiency. Maximum removal efficiency of approximately 93% was achieved under optimized conditions (pH 8, 60 min, and 40 °C). An ANN model was developed to capture the nonlinear relationships between operational parameters and adsorption performance. Model evaluation based on statistical indicators (R² and error metrics) demonstrated strong predictive capability and stable generalization behavior. The novelty of this work lies in integrating experimentally derived adsorption data with ANN-based predictive optimization for citrus peel biochar systems, providing a data-driven approach for improving adsorption process control beyond conventional empirical modeling. Overall, the findings highlight the potential of citrus peel biochar as a sustainable adsorbent and demonstrate the usefulness of ANN as a complementary predictive tool for adsorption process optimization.