Developing Qspr Models Of Gas/Particle Partition Coefficients Of Selected Persistent Organic Compounds (POPs)

Many persistent organic compounds(POPs)are semi-volatile,allowing them to volatilize into the atmosphere,which contain various particulates with different sizes.Because of the pore structures and organic components of the atmospheric particulates,POPs tend to adsorb on the surface of the particles or absorb into the organic matter,and then undergo transportation with the particulate movement and finally distribute between different environmental media.The partitioning of POPs between gas and particle phases observably affects their environmental migrations,transformation processes and environmental fates.Studying the distribution equilibrium of typical POPs between two phases could help us to understand their environmental fate and provide some basic data for ecological risk assessment.The gas/particle partition coefficient(Kp)is used to account for the partitioning equilibrium of compounds between gas phase and particle phase.However,experimental measurement of KP values for all semi-volatile organic chemicals is time-consuming,expensive and laborious,and is also limited by the increasing number of chenicals,making it is unpractical.Furthermore,the exsited KP predictive models are normally developed for single chemical species with small dataset size,and the mechanism explaination and predictive power also require improvement.Therefore,the purpose of this study is to collect the experimental KP values of POPs as many as possible and develop quantitative structure-property relationships(QSPR)models according to the guidance document stipulated by Organization for Economic Co-Operation and Development(OECD).Molecular mechanisms that govern partitioning equilibrium of chemicals between gas phase and particle phase were systematically alalyzed.The influence of ambient temperature on the KP values of chemicals was also accounted.(1)Poly-/perfluoroalkyl substances(PFASs)are a class of synthetic fluorinated organic substances that raise increasing concern because of their environmental persistence,bioaccumulation and widespread presence in various environment media and organisms.In this study,a temperature-dependent QSPR predictive model for log KP of PFASs was developed and the molecular mechanism that governs their partitioning equilibrium between gas phase and particle phase was analyzed.The regression model has a good statistical performance and robustness.The mechanism analysis indicates that electrostatic interaction and dispersion interaction play the most important role in the partitioning equilibrium.Based on the Williams plot,the compound 6:2 FTOH(T = 288.68 K)was diagnosed as an outlier.After removing it from the dataset,a new KP predictive model was rebuilded and its statistical performance has been significantly improved.The relationships between log KP and ambient temperature show that log KP values negatively correlate with ambient temperature except for three compounds 6:2 FTOH、MeFBSE and EtFOSE.High temperature results in large vapor pressure for these chemicals,consequently,easiness to become gaseous phase and decrease log KP.The developed QSPR model can be used to predict log KP values of neutral fluorotelomer alcohols and perfluor sulfonamides/sulfonamidoethanols with different substitutions at nitrogen atoms within the temperature variance(-4-16 ℃).(2)Polybrominated diphenyl ethers(PBDEs),typical brominated flame retardants(BFRs),have raised increasing attention due to their widespread manufacture,usage,disposal around the world and the frequent detection in a variety of environmental media.In the present study,the molecular mechanism partitioning equilibrium for PBDEs between gas and atmospheric particle phases was investigated,and a new temperature-dependent predictive model for the KP of PBDEs was developed.Molecular volume(V),the most negative atomic charge on a carbon atom(qC-),the most positive atomic charge on a hydrogen atom(qH+)and anbient temperature(T)were selected into the final model,indicating dispersion interaction,electrostatic interaction and hydrogen bond played the most important role in their partitioning between two phases.Most PBDEs obtained small Kp values under high environment temperature(negative correlation,R = 0.260-0.996),probably because of the high vapor pressure and large volatilization.The predictive model has good statistical performance,R2=0.852,RMSEP = 0.337,after removing the outlier(BDE-196,T= 304.65K).Validation results(R2 = 0.856,RMSEP = 0.348,MAE = 0.281,training set;R2 = 0.843,RMSEP = 0.316,MAE = 0.253,test set)further indicate the good predictive ability and robustness of the Kp predictive model.The application domain is defined as PBDE congeners with the number of substituted bromine atoms increasing from three(BDE-17,28)to ten(BDE-209)within the temperature variance(10-32 ℃).(3)Finally,an integral Kp predictive model for other four classes of chemicals including PAHs,PCBs,n-Alkanes and PCDD/Fs was established and the molecular mechanism of their partitioning equilibrium was also systematically investigated.The QSPR model showed good statistical performance and predictive robustness(R2 = 0.860,RMSEP = 0.335,MAE = 0.286,training set;R2 = 0.843,RMSEP = 0.326,MAE = 0.261,test set).Dispersion interaction,electrostatic interaction,hydrogen bond were the main intermolecular interactions which influence the distribution of chemicals between gas and atmospheric particle phases.In order to further study the factors that govern the partition equilibrium of chemicals,the QSPR model for each type of POPs was developed.The results showed that dispersion interaction is still the most important factor that affects the atmospheric distribution of four types POPs.For PCDD/Fs and PCBs,electrostatic interaction is another dominating factor.Ambient temperature is only screened into the Kp predictive models for PAHs and PCBs with a negative correlation relationship between log KP.