Synthetic bacterial consortium for degradation of plastic pyrolysis oil waste: experimental optimization and neural network modeling.

Introduction: The plastic crisis is omnipresent, ranging from plastic littering to microplastics found in every niche of the planet, including the human body. To achieve higher recycling quotas, especially for mixed plastic waste, pyrolysis is a viable option. However, plastic pyrolysis oil waste (PPOW) poses significant challenges for valorization due to its extreme molecular heterogeneity. Methods: To reduce this molecular heterogeneity, we artificially compounded, monitored, and optimized a bacterial consortium capable of tolerating organic pollutants and utilizing them as carbon and energy sources. Additionally, a back-propagation (BP) neural network was applied to evaluate O Results: The primary constituents of the PPOW were alkanes and ε-caprolactam. Within 7 days, the bacterial community demonstrated remarkable efficacy, degrading alkanes (C10-C28) at rates of 71-100% and achieving complete removal of ε-caprolactam (8,032 mg/mL) and naphthalene. The AI model predicted O Discussion: This AI-driven approach significantly reduces experimental iterations and operational costs for future process optimization. These results are discussed in the context of a developing open circular plastic economy.