基于QSAR关系预测PCBs和部分PAHs在LDPE中的扩散系数

分子扩散系数（D）是获得污染物与环境介质之间的平衡分配系数（K）的重要前提，然而通过实验测定获取污染物的扩散系数的过程过于繁琐，因此需开发一种更为简单、高效、准确的预测模型来定量预测扩散系数.为此，本文搜集了一些多环芳香烃（PAHs）和多氯联苯（PCBs）在低密度聚乙烯膜（LDPE）上扩散系数（log D）的实测值，基于定量结构-活性关系（QSAR），利用逐步多元线性回归（MLR）构建了预测D值的模型.模型的决定系数R_adj²为0.941，交叉验证系数Q_LOO²为0.934，外部系数Q_ext²为0.895.结果表明，该QSAR模型具有良好的拟合优度、稳健性和预测能力，其可用来预测应用域内有机污染物在LDPE膜上的扩散系数.

Development of QSAR model for predicting diffusion coefficients of PCBs and PAHs in LDPE

The application of low density polyethylene (LDPE) as passive sampling devices for monitoring the concentration of hydrophobic organic contaminants (HOCs) requires data on diffusion coefficients (D) for the estimation of uptake rates. Most of the diffusion coefficients are usually obtained from experimental measurements, which are not readily available for all potential pollutants. Therefore, current work aimed to establish mathematical models for predicting D values with the physicochemical properties of chemicals. To make further improvements in measuring D values, this study focused to develop a quantitative structure-activity relationship (QSAR) model for predicting diffusion coefficients. The results of stepwise multiple regression indicated that QSAR model fits well with objectives, and had robustness and predictive capacity, with the determination coefficients (R²_adj) of 0.941, cross-validation coefficients (Q²_LOO) with 0.943, and with external validation coefficient (Q²_ext) of 0.895. Mechanism interpretation suggested that the main factors governing the diffusion process in LDPE were van der Waals volumes. The results of current study provide an excellent tool for predicting D values of HOCs within the applicability domains.