Optimizing biodegradation of polychlorinated biphenyls (PCBs) using bacteria isolated from common effluent treatment plant (CETP) sludge: Integration of machine learning, kinetic studies, and metabolomic analysis

Abstract

The present study demonstrates the effective biodegradation of six polychlorinated biphenyls (PCBs) using microbial consortia consisting of Bacillus sp. and Stutzerimonas sp. Isolated from common effluent treatment plant (CETP) sludge for sustainable environmental remediation. The process during the degradation was optimized using machine learning Artificial Neural Networks (ANN) and Response Surface Methodology (RSM) with Box-Behnken Design (BBD), leading to high degradation efficiencies: 100% for PCB28 and PCB52, 98.89% for PCB101, 98.85% for PCB138, 95.89% for PCB153, and 88.56% for PCB180. The metabolites generated during the study are less toxic and nontoxic. The consortium followed two degradation pathways; acetyl CoA and Pyruvic acid. Further, the degradation kinetics studied during the process reveals that PCB138,153and 180 followed first-order kinetics, while PCB 28, 52, and 101 followed second-order kinetics. Moreover, the bacterial growth kinetics were studied using the Gompertz fitting curve with an R² value of 0.9405. The findings underscore the importance of fine-tuning biodegradation processes to enhance their efficiency, ultimately contributing to greener, more sustainable methods for PCB remediation. By integrating machine learning and statistical optimization, this study paves the way for future advancements in environmental biotechnology aimed at mitigating the adverse impacts of PCBs on ecosystems and human health.