Department of Chemistry, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo, 113-0033, Japan.
Angew Chem Int Ed Engl. 2023 Mar 27;62(14):e202219059. doi: 10.1002/anie.202219059. Epub 2023 Feb 23.
"How do we decide the stoichiometry of host-guest complexes?" This question has long been answered by the Job plot since its first report in 1928. However, as the Job plot was claimed to be misleading in 2016, the question became an open question again and called for renewed investigations. An information-theoretic approach, called Akaike's information criterion, is introduced in this study to select the best model of host-guest complexes, which can rank the models with weight of evidence. A few test cases with unique cylindrical hosts were examined to demonstrate the applicability of the information-theoretic method. Consequently, reasonable views over the thermodynamic behaviors of dumbbell-and-cylinder complexes were obtained. Akaike's information criterion can be a useful and superior alternative to statistical null hypothesis testing, which was proposed as a remedy in place of the Job plot.
“我们如何确定主客体配合物的化学计量比?”自 1928 年首次报道以来,这个问题一直由 Job 图来回答。然而,由于 2016 年有人声称 Job 图具有误导性,这个问题再次成为一个悬而未决的问题,需要重新研究。本研究引入了一种信息论方法,称为赤池信息量准则,用于选择最佳的主客体配合物模型,该模型可以用证据权重对模型进行排序。本文通过几个具有独特圆柱主体的测试案例来验证信息论方法的适用性。结果,获得了关于哑铃-圆柱配合物热力学行为的合理观点。赤池信息量准则可以作为统计零假设检验的有用且优越的替代方法,后者被提议作为替代 Job 图的方法。