基于TLSER关系预测POPs在PUF与空气中的分配系数

为开发一种简洁高效预测K_PA值的方法，本文收集了包含多氯联苯（PCBs）、苯类（Benzenes）、多环芳香烃（PAHs）和杀虫剂（Pesticides）等，共95种污染物在聚氨酯泡沫塑料（PUF）与空气中分配系数（K_PA）的实验值，基于理论线性溶解能关系（TLSER），运用逐步多元线性回归（MLR）构建了预测K_PA值的模型.结果表明，模型的决定系数R_adj²为0.901，外部验证系数Q_ext²为0.748.TLSER模型有较好的拟合度、稳健性和预测能力，可以预测应用域内POPs在PUF和空气之间的分配系数.

Development of TLSER model for prediction partition coefficients of POPs between polyurethane foam and air

This work developed a theoretical linear solvation energy relationship (TLSER) model to predict the partition coefficient between polyurethane foam (PUF) and air (K_PA). The date set of TLSER model covered 95 compounds comprising polychlorinated biphenyls (PCBs), Benzenes, polycyclic aromatic hydrocarbons (PAHs) and Pesticides. A multiple variate linear model was then used to estimate the partition coefficient. The results showed that the adjusted determination coefficient (R²adj) was 0.901, and the external validation coefficient (Q²ext) was 0.748, which indicated that the TLSER model had satisfactory goodness-of-fit, robustness and good predictability. Consequently, our results suggested the proposed model could be used to predict the partition coefficients of other POPs within the application domain between PUF and air.