芳香族化合物与水合电子水相反应速率常数的QSAR模型研究

QSAR models for predicting the aqueous reaction rate constants of aromatic compounds with hydrated electrons

基于水合电子(eaq^-)的高级还原技术(ARPs)是一种有效地去除水中痕量有机污染物的技术.有机物与水合电子反应的速率常数(keaq-)是评价高级还原处理体系中有机物去除效率的一个重要参数.然而化学品种类繁多,通过实验方法一一获取keaq-是不现实的.因此,需要建立一种能够快速预测keaq-的方法,填补现有数据的缺失.本研究搜集整理了94种芳香族化合物的keaq-实验值,采用逐步多元线性回归(MLR)和支持向量机(SVM)方法分别构建了预测化合物的keaq-线性和非线性定量结构活性关系(QSAR)模型.两个模型均具有良好的拟合度(Radj,tr^2> 0.800)、稳健性(Qext^2=0.782)和预测能力(Qext^2> 0.790).机理分析表明,最低未占据分子轨道能(ELUMO)和偶极矩加权的来自扩增边缘邻接处的最大特征值(SpMADAEA(dm))是影响有机物同eaq^-反应活性的重要参数,此外,keaq-还与化合物的键级和极化率有关.元分析结果表明具有吸电子官能团的芳香化合物比含供电子基团的化合物倾向于与eaq^-有更高的反应活性.

Hydrated electron(eaq^-)-based reduction processes are promising for degrading organic pollutants in water engineering systems. The second order rate constant(keaq-) of eaq- with organic compounds is an important parameter for evaluating the removal efficiency of organic pollutants in advanced reduction proless. However, experimental determination of keaq-seems fairly unrealistic because of the large number of organic chemicals. Thus, it is necessary to develop an effective method to predict keaq-. In this study, the experimental keaq-values of 94 aromatic compounds were collected. The quantitative structure-activity relationship(QSAR) models for predicting keaq-were constructed by stepwise multiple linear regression(MLR) and support vector machines(SVM) methods, respectively. Both two models had satisfactory goodness-of-fit(Radj,tr^2 > 0.800), robustness(Qext^2= 0.782) and good predictability(Qext^2> 0.790). Mechanistic analysis reveald that the energy of the lowest unoccupied molecular orbital(ELUMO) and the spectral mean absolute deviation from augmented edge adjacency mat weighted by dipole moment(SpMADAEA(dm)) were the most important descriptors. Additionally, keaq- was found to be related to the bond order and polarizability of the compounds. The meta-analysis showed that aromatic compounds with electron-withdrawing functional groups tended to have higher reactivity with eaq^- than those containing electron-donating groups.