摘要
应用物质结构-性质关系(QSPR)方法研究了34种羧酸pKa与分子描述符之间的定量关系,采用DRAGON 5.4软件和Hyer Chem7计算了34种羧酸分子的568种结构类参数和12种量子化学参数。根据偏差权重法筛选出61种结构描述符,采用多元逐步回归分析建立了34种羧酸结构-性质关系,得到最佳方程。方程相关系数R=0.990,标准偏差S=0.170。且有F>Fα,表明方程回归效果显著;对回归系数的t检验结果均有Sig<0.05,表明6个参数对pKa影响显著。"逐一剔除"交叉验证的结果证明模型具有良好的稳定性和较强的预测能力。6个参数物理意义及其对pKa影响的深入分析表明,其中羧基中氢原子净电荷(QH)和羧基中非羟基氧原子净电荷(Q=O)是影响羧酸强度的关键因素。
Quantitative structure-property relationship （QSPR） method is performed to study the relationship between Pk values and molecular descriptors of 34 sorts of substituted carboxylic acids. With the DRAGON 5.4 software and HyerChem 7, 568 structural parameters and 12 quantum chemistry parameters for 34 sorts of substituted carboxylic acids are calculated. 61 structural descriptors are chosen by using the variable selection method of weighted deviation. Through stepwise regress, a equation of QSAR of 34 sorts of substituted carboxylic acids has been successfully proposed. The R （correlation） and S （standard error） are O. 990 and 0. 170, respectively. Furthermore, the two-sided t-test was used to validate the significance of the 6 descriptors. With the statistically significance （ P〈 0. 05 ） , correlation coefficients of descriptors show that the 6 descriptors have important impaction the pKa values. Furthermore, leave-one- out cross validation as performed, the results confirmed the good stability and powerful prediction of this model. The physical meaning and the influence of the 6 parameters on pK of substituted carboxylic acids were investigated. The two parameters, net charge of the hydrogen atoms in carboxyl （QH） and net charge of the oxygen atom in non-hydroxyl （ Q=0） are the key factors affecting the acid intensity of substituted carboxylic acids.
出处
《实验室研究与探索》
CAS
北大核心
2015年第2期20-23,33,共5页
Research and Exploration In Laboratory
基金
国家教育部春晖计划(Z2012108)
