Fu Jia, Li Meng, Rong Chunying, Zhao Dongbo, Liu Shubin
College of Chemistry and Chemical Engineering, Hunan Normal University, Changsha, Hunan, 410081, People's Republic of China.
Institute of Biomedical Research, Yunnan University, Kunming, 650500, People's Republic of China.
J Mol Model. 2024 Sep 18;30(10):341. doi: 10.1007/s00894-024-06116-7.
Electrophilicity and nucleophilicity are two vastly important chemical concepts gauging the capability of atoms in molecules to accept and donate the maximal number of electrons. In our earlier studies, we proposed to simultaneously quantify them using the Kullback-Leibler divergence from the information-theoretic approach in density functional theory. However, several issues with this scheme remain to be clarified such as its general validity, predictability, and relationship with other information-theoretic quantities. In this work, we revisit the matter with bigger datasets and deeper theoretical insights. Five information-theoretic quantities including Kullback-Leibler divergence, Hirshfeld charge, Ghost-Berkowitz-Parr entropy, and second and third orders of relative Onicescu information energy are found to be reliable and robust descriptors of electrophilicity and nucleophilicity propensities. Employing these five descriptors, we design a list of new compounds and predict their electrophilicity and nucleophilicity scales. This work should markedly improve our confidence and capability in applying information-theoretic quantities to evaluate electrophilicity and nucleophilicity propensities and henceforth pave the route for more applications of these quantities from information-theoretic approach in density functional theory in the future.
All structures were fully optimized at the M06-2X/6-311 + G(d) level of DFT functional using the Gaussian 16 package (version C01) with integration grids and tight self-consistent-field convergence. The solvent effect was taken into account by using the implicit solvent model (CPCM) in the CHCl solvent, and all 3D contour surfaces of Fukui function, local temperature, and ITA (information-theoretic approach) quantities were generated by GaussView. The Multiwfn 3.8 program was used to calculate the ITA indexes and atomic charges.
亲电性和亲核性是两个极为重要的化学概念，用于衡量分子中原子接受和给出最大电子数的能力。在我们早期的研究中，我们提议使用密度泛函理论中信息论方法的库尔贝克-莱布勒散度来同时对它们进行量化。然而，该方案的几个问题仍有待阐明，例如其普遍有效性、可预测性以及与其他信息论量的关系。在这项工作中，我们使用更大的数据集和更深入的理论见解重新审视了这个问题。发现包括库尔贝克-莱布勒散度、赫希菲尔德电荷、幽灵-伯科维茨-帕尔熵以及相对奥尼塞库信息能量的二阶和三阶在内的五个信息论量是亲电性和亲核性倾向的可靠且稳健的描述符。利用这五个描述符，我们设计了一系列新化合物，并预测了它们的亲电性和亲核性标度。这项工作应显著提高我们应用信息论量来评估亲电性和亲核性倾向的信心和能力，从而为未来在密度泛函理论中从信息论方法更多地应用这些量铺平道路。
使用高斯16软件包（版本C01），在DFT泛函的M06 - 2X/6 - 311 + G(d)水平下，采用积分网格和紧密的自洽场收敛对所有结构进行了完全优化。通过在CHCl溶剂中使用隐式溶剂模型（CPCM）来考虑溶剂效应，并且通过GaussView生成福井函数、局部温度和ITA（信息论方法）量的所有三维等值面。使用Multiwfn 3.8程序计算ITA指数和原子电荷。
