An error was noticed in Table 3 of the original article after publication. In the determination of the activation energies of the dielectric relaxations, a numerical mistake was made by not properly considering that the latter are defined in terms of a natural logarithm while the experimental data are represented in a base-ten logarithmic scale. The mistake affected the activation energy of the secondary relaxations and the determination of the fragility strength coefficient D and kinetic fragility index m_p associated with the primary relaxation. As a consequence of the logarithmic base change, several entries of Table 3 need to be corrected by a factor ln(10). We have replaced the faulty table with the new Table 3 reported here below. With this correction, the fragility of each compound is now coherent with the effective activation energy of the primary (α) relaxation at T_g, which was computed correctly in the original paper. In the abstract, we have correspondingly replaced the value of the kinetic fragility index (which was erroneously given as m_p ≈ 32, a relatively low value) with the correct value of m_p ≈ 73, which is a more common value for molecular glass formers. We have done the same replacement at the end of the Discussion section. This numerical correction does not change the main results and conclusions of the original article. In particular, it remains true that the fragility indices are virtually identical in all three diazepine derivatives, and therefore that this parameter cannot be employed as a reliable predictor of the crystallization tendency. It is worthwhile to point out that this numerical correction further corroborates one important conclusion of our work, namely, the molecular interpretation of the secondary γ relaxation process. In fact, the correct activation energies of the γ relaxations of diazepam and of its nordazepam derivative are 72 ± 9 and 58 ± 5 kJ/mol, respectively, which are now compatible with those found in previous studies by NMR experiments and ab initio calculations, where the conformational ring-inversion activation energies were reported to be 74 and 52 kJ/mol(1) and in the ranges 72.4–74.1 and 44.8–47.3 kJ/mol,(2) respectively, for the two compounds. This agreement confirms the assignment of the γ relaxation to the diazepine ring-inversion dynamics. On page 1827, right column, we have correspondingly modified the sentence in which we reported the discrepancy between the activation energy of the γ relaxations of Table 3 with those of previous studies in solution, and removed the rationalization that we had given in terms of a possible difference between the liquid and glassy state. As a matter of fact, our correction allows reaching the further, interesting conclusion that the activated behavior of the ring-inversion relaxation of liquid diazepines is not affected by the transition to the glass state. We thank Prof. H.P. Diogo from the Chemistry Engineering Department of the Universidade Técnica de Lisboa for first observing the incongruency between the activation energy and the fragility of the primary relaxation that has led us to check our results. This article references 2 other publications.

中文翻译：

---

更正“三种药用活性苯二氮卓衍生物的比较物理研究：结晶与无定形状态和结晶趋势”

发表后发现原文表3有错误。在确定介电弛豫的活化能时，由于没有正确考虑后者是根据自然对数定义的，而实验数据以十对数为底的对数标度表示，因此造成了数值错误。该错误影响了二次弛豫的活化能和脆性强度系数D和动脆性指数m _p的确定与初级松弛有关。作为对数基数变化的结果，表 3 中的几个条目需要通过因子 ln(10) 进行校正。我们已经用下面报告的新表 3 替换了故障表。通过这种校正，每种化合物的脆性现在与T _g处初级 (α) 弛豫的有效活化能一致，这在原始论文中已正确计算。在摘要中，我们相应地用正确的 m p 值替换了动力学脆性指数的值（它被错误地给出为m _p ≈ ₃₂，一个相对较低的值）≈ 73，这是分子玻璃形成剂更常见的值。我们在讨论部分的最后做了同样的替换。本次数值修正不改变原文章的主要结果和结论。特别是，所有三种二氮杂卓衍生物的脆性指数几乎相同，因此该参数不能用作结晶趋势的可靠预测指标。值得指出的是，这一数值修正进一步证实了我们工作的一个重要结论，即二次γ弛豫过程的分子解释。事实上，地西泮及其去甲西泮衍生物的 γ 弛豫的正确活化能分别为 72 ± 9 和 58 ± 5 kJ/mol，从头算计算结果表明，这两种化合物的构象环反转活化能分别为 74 和 52 kJ/mol(1)，范围分别为 72.4–74.1 和 44.8–47.3 kJ/mol,(2)。该协议证实了 γ 弛豫与二氮杂环反转动力学的关系。在第 1827 页，右栏，我们相应地修改了我们报告表 3 的 γ 弛豫的活化能与先前在溶液中研究的活化能之间的差异的句子，并删除了我们根据 a 给出的合理化液态和玻璃态之间的可能差异。事实上，我们的修正允许更进一步，有趣的结论是液体二氮杂环的倒环弛豫的激活行为不受向玻璃态转变的影响。我们感谢里斯本科技大学化学工程系的 HP Diogo 教授首次观察到活化能与初级弛豫的脆弱性之间的不一致，这使我们检查了我们的结果。本文引用了其他 2 篇出版物。