High-throughput prediction of fabric-air partition and diffusion coefficients for indoor organic compounds: A QSPR modeling framework.

Fabric materials serve as both significant sinks and secondary emission sources for organic pollutants in indoor environments. Accurately determining their mass-transfer parameters is essential for assessing indoor pollutant fate and exposure risk. However, existing measurement methods are time-consuming and costly, and available predictive models are limited to single parameters, constraining prediction efficiency and accuracy. This study develops a dual-parameter QSPR framework for the unified prediction of the apparent diffusion coefficient (Dm) and fabric-air partition coefficient (Kfa) of organic compounds in indoor fabrics. By embedding porous-media mass-transfer mechanisms into the model structure and replacing experimentally determined material parameters with DFT-derived electronic descriptors, most model inputs become computable from molecular. This makes the framework free of any compound-specific fabric sorption measurements. The reliability of the dual-parameter framework is confirmed through internal and external validation against literature data and independent sandwich chamber experiments. Both models achieve R2adj and Q2LOO above 0.94, with chamber-validation residuals within ±1. This study substantially improves prediction efficiency, supporting high-throughput screening of indoor organic pollutant fate, exposure risk assessment, and indoor air quality management. The approach also offers a transferable methodology for rapid mass-transfer prediction in other organic compound-porous material systems.