Bioremediation of a Dis-Azo RB222 Dye using a Novel Bacillus Strain: Experimental and Artificial Neural Network Based Process Optimization

Abstract

Azo dyes, predominantly used in textile industries, pose significant hazards to aquatic ecosystems due to their recalcitrant and toxic nature. Although bioremediation has been explored for the degradation of various azo dyes, the biological removal of “dis-azo” dyes such as Reactive Blue 222 (RB222) remains limited. Moreover, there are no documented instances of using a Bacillus strain to remove RB222 in either batch or continuous bioreactor systems. This study demonstrated the bioremediation of a dis-azo dye, RB222, using a novel Bacillus strain, and optimized process conditions for their efficient degradation in both batch and continuous bioreactor systems. Specifically, a new bacterial strain, Bacillus subtilis MN372379, isolated from a soil sample at a dye-discharging site in Bhadohi district, Uttar Pradesh, India, was immobilized on the waste polyurethane foam scrap to degrade RB222 in both batch and continuous packed bed bioreactors (PBBR). In the batch bioreactor, the performance of the immobilized biomass was compared with a free cell system of suspended biomass, and the optimal parameters for process duration, pH level, temperature, and glucose concentration were determined to be 6.0 days, 6.5, 30°C, and 1 g/L, respectively. Under the optimized conditions, the removal efficiencies (REs) of RB222 for the free and immobilized batch bioreactors were obtained as 87.94 and 95.32%, respectively, for the initial dye concentration of 50 ppm. The RE of RB222 was further enhanced to 98.56% in a continuously operated PBBR at an initial dye concentration of 50 mg/L and an inlet loading rate (ILR) of 30 mg/L·day, highlighting the RB222-degrading capability of the isolated bacterial strain. Finally, an artificial neural network (ANN) model was developed using the Levenberg–Marquardt algorithm to predict the REs of the RB222 dye in both free and immobilized cell systems. The model demonstrated excellent predictive accuracy, with regression coefficients of 0.89 and 0.92 for the free and immobilized cell systems respectively, indicating a strong correlation between the predicted and measured values. The study highlighted the potential of Bacillus subtilis MN372379 for the bioremediation of RB222-containing wastewater. Future research could explore the identification of additional bacterial strains capable of degrading recalcitrant dis-azo dyes, with the goal of developing a microbial consortium to improve bioremediation performance in a scaled-up PBBR.