Department of Chemistry, Università degli Studi di Torino, via Pietro Giuria 7, 10125, Torino, Italy.
Institute of Inorganic and Analytical Chemistry, Goethe University, Max-von-Laue-Strasse 7, 60438, Frankfurt am Main, Germany.
Chemistry. 2022 Jan 27;28(6):e202103589. doi: 10.1002/chem.202103589. Epub 2021 Dec 28.
Among all possible NMR crystallography approaches for crystal-structure determination, crystal structure prediction - NMR crystallography (CSP-NMRX) has recently turned out to be a powerful method. In the latter, the original procedure exploited solid-state NMR (SSNMR) information during the final steps of the prediction. In particular, it used the comparison of computed and experimental chemical shifts for the selection of the correct crystal packing. Still, the prediction procedure, generally carried out with DFT methods, may require important computational resources and be quite time-consuming, especially if there are no available constraints to use at the initial stage. Herein, the successful application of this combined prediction method, which exploits NMR information also in the input step to reduce the search space of the predictive algorithm, is presented. Herein, this method was applied on mebendazole, which is characterized by desmotropism. The use of SSNMR data as constraints for the selection of the right tautomer and the determination of the number of independent molecules in the unit cell led to a considerably faster process, reducing the number of calculations to be performed. In this way, the crystal packing was successfully predicted for the three known phases of mebendazole. To evaluate the quality of the predicted structures, these were compared to the experimental ones. The crystal structure of phase B of mebendazole, in particular, was determined de novo by powder diffraction and is presented for the first time in this paper.
在所有可能的 NMR 晶体学方法中，用于确定晶体结构的方法是晶体结构预测-NMR 晶体学（CSP-NMRX）。最近，它已成为一种强大的方法。在后者中，原始程序在预测的最后步骤中利用固态 NMR（SSNMR）信息。特别是，它使用计算和实验化学位移的比较来选择正确的晶体堆积。然而，预测过程通常使用 DFT 方法进行，可能需要重要的计算资源并且非常耗时，特别是如果在初始阶段没有可用的约束条件可供使用。在此，介绍了成功应用这种联合预测方法的情况，该方法在输入步骤中利用 NMR 信息来减少预测算法的搜索空间。在此，该方法应用于具有变构现象的甲苯咪唑。将 SSNMR 数据用作选择正确互变异构体和确定晶胞中独立分子数的约束条件，可大大加快进程，减少要执行的计算数量。通过这种方式，成功预测了甲苯咪唑的三种已知相的晶体堆积。为了评估预测结构的质量，将它们与实验结构进行了比较。甲苯咪唑 B 相的晶体结构是通过粉末衍射从头确定的，本文首次提出。
