Bosque Ramón, Sales Joaquim
Departament de Química Inorgànica, Universitat de Barcelona, Martí i Franquès, 1, 08028-Barcelona, Spain.
J Chem Inf Comput Sci. 2003 Mar-Apr;43(2):637-42. doi: 10.1021/ci025632e.
A Quantitative Structure-Property Relationship (QSPR) is developed for the O-H bond dissociation energy (BDE) of a set of 78 phenols. The data set was composed of monosubstituted, disubstituted, and polysubstituted phenolic derivatives containing substituents with different steric and electronic effects in the ortho-, meta-, and para-positions of the aromatic ring. The proposed model, derived from multiple linear regression, contains seven descriptors calculated solely from the molecular structure of compounds. The average absolute relative errors are 1.37% (R(2) = 0.8978; SD: 6.67) and 1.13% (R(2) = 0.9076; SD: 4.26) for the working set (62 compounds) and the prediction set (16 compounds), respectively. These results are better than those obtained from DFT calculations, QSAR approach, and correlations with Hammet parameters.
针对一组78种酚类化合物的O-H键离解能（BDE），建立了定量结构-性质关系（QSPR）。数据集由单取代、双取代和多取代酚类衍生物组成，这些衍生物在芳环的邻位、间位和对位含有具有不同空间和电子效应的取代基。从多元线性回归得出的所提出模型包含仅根据化合物分子结构计算出的七个描述符。工作集（62种化合物）和预测集（16种化合物）的平均绝对相对误差分别为1.37%（R(2) = 0.8978；标准差：6.67）和1.13%（R(2) = 0.9076；标准差：4.26）。这些结果优于从密度泛函理论（DFT）计算、定量构效关系（QSAR）方法以及与哈米特参数的相关性所获得的结果。
