The orientation indices and group corresponding indices of 210 molecular structures of polychlorinated biphenyls(PCBs)and biphenyl were calculated.By combining these indices with seven thermodynamic properties reporte...The orientation indices and group corresponding indices of 210 molecular structures of polychlorinated biphenyls(PCBs)and biphenyl were calculated.By combining these indices with seven thermodynamic properties reported in literature of such PCBs, including standard entropy (S0),standard enthalpy(H0),standard free energy(G0),total energy(ET),thermal energy correction(Eth),zero-point vibration energy(Ezpv)and constant volume molar heat capacity (C0V), a series of quantitative structure property relationship (QSPR) equations were built up by the multiple regression method.All the correlation coefficients were larger than 0.99.These models could better explain the change of thermodynamic properties for the PCBs,and the QSPR models had high correlation coefficients,good stability and good predictability.A modified leave-one-out procedure test was performed to validate model robustness,and predict the S0,H0,G0,ET,Eth,Ezpv,C0V of other PCBs molecules by the regression equation,and the error analysis were discussed.The results showed better agreement between the values predicted by the model developed and the observed values of the thermodynamic properties.展开更多
文摘The orientation indices and group corresponding indices of 210 molecular structures of polychlorinated biphenyls(PCBs)and biphenyl were calculated.By combining these indices with seven thermodynamic properties reported in literature of such PCBs, including standard entropy (S0),standard enthalpy(H0),standard free energy(G0),total energy(ET),thermal energy correction(Eth),zero-point vibration energy(Ezpv)and constant volume molar heat capacity (C0V), a series of quantitative structure property relationship (QSPR) equations were built up by the multiple regression method.All the correlation coefficients were larger than 0.99.These models could better explain the change of thermodynamic properties for the PCBs,and the QSPR models had high correlation coefficients,good stability and good predictability.A modified leave-one-out procedure test was performed to validate model robustness,and predict the S0,H0,G0,ET,Eth,Ezpv,C0V of other PCBs molecules by the regression equation,and the error analysis were discussed.The results showed better agreement between the values predicted by the model developed and the observed values of the thermodynamic properties.
