阴离子表面活性剂临界胶束浓度的QSPR研究

Quantitative Structure-Property Relationship of the Critical Micell Concentration of Anionic Surfactant

以价分子连接指数(0χ,1χ,2χ)为自变量,临界胶束浓度(cmc)为因变量,采用QSPR方法和SPSS 21.0统计软件建立了样本数为40的数学模型。最佳模型为lg cmcobs=2.245-0.7252χ,复相关系数R2=0.864。从预测方程可看出临界胶束浓度与2χ呈负相关关系。筛选过程中发现,0χ,1χ与函数值之间线性关系微弱,与2χ共线性较强,确定了临界胶束浓度最重要影响因素为2阶价分子连接指数(2χ)。内部验证使用留一法(Leave-one-out),得到的交叉验证系数q2=0.831,证明模型稳定性和预测性良好。残差频率分布直方图基本符合正态分布,实验预测值对比图中大部分样本点落在对角线附近,具有明显线性关系。从统计量及图表可证明,该方程可以用来预测阴离子表面活性剂的临界胶束浓度。

A 40-sample mathematical model derived from QSPR is built using Enter Method of SPSS 21.0with molecular connectivityindex of rank zero to two as the independent variable and critical micell concentration as dependent variable.The best model is whose multiple correlation coefficients.It suggests that is negative correlation with the.The weak linear relationship between collining with strongly and function value is found in the screening process,so that the most important factor of is is made sure.Leave-one-out method is chosen for internal verification as a result of q2=0.831,confirming the good stability and predictability.Residuals frequency distribution histogram mainly conform to normal distribution,and with obvious linear relationship sample point in contrast diagram of experimental value and predicted value.It＇s proved that this equation is able to predict the values of cationic surfactant through the statistics and statistical chart.