拓扑邻接电负性指数法预测多氯代联苯醚(PCDEs)的理化性质

Topological Adjacent Electronegative Index for Prediction of the Physicochemical Property of PCDEs

将PCDEs分子中苯环上的原子或基团(C、O、CH和CCI)作为分子碎片,构建分子的邻接矩阵．再以相应原子或基团的均衡电负性构建行矩阵．结合行矩阵和邻接矩阵,计算得到一个新的指数??拓扑邻接电负性指数(TAEI)．以107个PCDEs的理化性质实验值为建模样本,用线性回归方法,建立PCDEs的饱和蒸气压(P_L^0)、水溶解度(S_(W．L)和正辛醇/水分配系数(K_(OW))的定量结构??性质相关(QSPR)模型,有良好的相关性(R＞0．97)．特别是对饱和蒸气压(P_L^0),获得了比已报道的量子化学方法更好的结果．用LOO法检验模型的稳定性和预测能力,所得交互检验的相关系数均在0．95以上．结果表明所建模型复相关系数高、稳定性好、预测能力强．图4,表1,参10．

The atoms or functional groups（C, O, CH and CCl） were used as molecular fragments to build an adjacent matrix, and the balanced electronegative of the molecular fragment was used to build a row matrix. The adjacent matrix and row matrix were combined to build a new index - topological adjacent electronegative index （TAEI）. The TAEI was combined with three observed physicochemical properties （vapor pressure, aqueous solubility and n-octanol/water partition coefficient） and three good quantitative structure-property relationship models （R〉0.97） were got by unitary linear regression method. One of the three models was better than the quantum chemical method. In order to test the stability and prediction ability of each model, the LOO cross validation was performed good results （Rscv〉0.95）were obtained. These results show that those models have high relative coefficients, good stability and good predictability. 4figs., 1tab., 10refs.