Insights into the dissipation and absorption behavior of Benzotriazole ultraviolet stabilizers in soil-plant systems: Experiments, molecular simulations, and machine learning.

Benzotriazole ultraviolet stabilizers (BUVs) are widespread emerging contaminants detected in soil-plant systems. However, the soil dissipation behavior and root absorption mechanisms of structurally diverse BUVs remain unclear. This study introduces a comprehensive framework integrating exposure experiments, molecular simulations, and machine learning (ML) to address these issues. We found that van der Waals forces drove a hydrophobic distribution mechanism and the coordination of the molecular spatial configuration primarily influenced root absorption of BUVs. Low-molecular-weight (MW) BUVs exhibited stronger binding to lipid transfer proteins (LTPs) due to favorable electrostatic potential differences, whereas high-MW BUVs faced steric hindrance, reducing their affinity for LTPs. Chlorine atoms decreased entropy loss by increasing local negative potential, and phenolic cyclic para-alkyl substituents enhanced LTP binding by expanding the neutral potential region. Furthermore, multi-scale ML models were developed to predict the root concentration factor (RCF), integrating molecular features, soil properties, and experimental conditions. The analyses were validated, and an optimized rhizosphere reinforcement strategy was proposed. The current study presents an innovative closed-loop paradigm, transitioning from experiments to molecular simulations and multi-scale ML predictions within the field of plant absorption. The findings provide theoretical support for phytoremediation technologies and structural insights into green additive alternatives.