The high water solubility and long-term thermal and optical stability of Reactive Blue 248 pose a challenge for its removal from industrial wastewater. The performance of BiOI as a photocatalyst under various conditions was investigated. Additionally, machine learning models were used to model and predict the efficiency of pollutant removal under different conditions. Furthermore, the stability and reusability of BiOI were evaluated for repeated cycles of dye degradation. Organic dye oxidation and identification of active radical species were evaluated through chemical oxygen demand (COD) and radical scavenging tests, respectively. The catalyst dose and contact time significantly influence dye removal efficiency, with p-values of 0.0299 and 0.0238, respectively. The optimal conditions include a pH of 6.713, a contact time of 60 min, and a catalyst dose of 0.634 g/L. The addition of AgNO3 as an e− scavenger had no effect, keeping the efficiency at 73.87%, indicating that free electrons do not play a significant role in the degradation of RB248. However, the introduction of KI, which scavenges h+, led to a substantial drop in efficiency to 40.27%, confirming the crucial role of h+ in the photocatalytic process. The second-order kinetic model provides a more accurate description of the dye removal process. The degradation process (COD Tests) was evaluated for three initial dye concentrations of 20, 30, and 50 mg/L. These results indicate that the degradation efficiency was higher at lower initial concentrations, while the removal process was slower at higher concentrations. Among the machine learning models evaluated on the test data, the ERT model outperforms others in all key performance metrics.
Makhdoomi et al. (Sun,) studied this question.