Comparative effectiveness of PDMS passive sampling versus Tenax-TA extraction for assessing bioavailable SVOCs in...

This study investigated the bioavailability of typical semi-volatile organic compounds (SVOCs), including organophosphate esters (OPEs), phthalates (PAEs), and polycyclic aromatic hydrocarbons (PAHs), in paddy and upland rice soils across their life cycle using polydimethylsiloxane (PDMS) passive sampling and Tenax-TA® sequential extraction. SVOC concentrations were significantly higher in paddy soils than in upland soils, primarily due to irrigation inputs. However, contamination levels in rice tissues were comparable between the two systems, possibly due to the similar PDMS-measured bioavailable concentrations in both soils. Comparative analysis showed that PDMS sampling correlated more strongly with SVOC accumulation in roots and provided regression slopes closer to unity than Tenax extraction. PDMS sampling effectively reflected SVOC dynamics in the rhizosphere and predicted grain uptake, with accuracy improving towards grain maturity. PDMS outperformed the Tenax method by more accurately reflecting the freely dissolved bioaccessible fraction. Integration of experimental data with multiple linear regression and machine learning models (Random Forest, Extra Trees, XGBoost) allowed the identification of key drivers for plant uptake, including bioavailable concentration, biodegradation half-life, and molecular descriptors. The results confirm PDMS sampling as a reliable and field-practical tool for the long-term in-situ evaluation of SVOC bioavailability, providing a mechanistic basis for establishing bioavailability-driven food safety strategies in complex rice agroecosystems.