Predicting the Gas Chromatographic Relative Retention Time of Polybrominated Diphenyl Ethers by MEDV-13 Descriptors

Abstract

Polybrominated diphenyl ethers (PBDEs) have been used extensively over the past two decades as flame or fire retardants added into many products such as textiles, electronic equipment and insulation material. PBDEs are now ubiquitous environmental pollutants and have been detected in biotic and abiotic matrices that include fish, birds, sediments, air, marine mammals, and human plasma and milk. The purpose of the paper is to develop some models which can predict the gas chromatographic relative retention time of PBDEs on seven stationary phases. The molecular electronegativity distance vector based on 13 atomic types (MEDV-13) were employed to describe the structural changes of PBDEs. Using the variables selection and modeling based on the prediction (VSMP), some four-variable linear equations from 80 PBDEs in the training sets with the correlation coefficient (r2) >0.98 and the root mean square errors of estimation <0.029 (except for CP-Sil19) were established. The standard regression coefficient indicated that MEDV descriptor x91 (interaction between the substructure –Br and –Br) was the most important descriptor to affect the RRT values of PBDEs. At the same time, it was found that the number of the ortho-substituent bromines and the number of bromine atoms in PBDE molecule may improve the predictive quality.