As part of a continuing project, the challenging room-temperature crystal structures of eight commercial pharmaceutical APIs have been solved by Monte Carlo simulated annealing techniques using synchrotron X-ray powder diffraction data (11-BM at APS), and optimized using density functional techniques. Tofacitinib dihydrogen citrate (Xeljanz®), (C15H21N6O)(H2C6H5O7), crystallizes in P212121 with a = 5.91113(1), b = 12.93131(3), c = 30.43499(7) Å, V = 2326.411(6) Å3, and Z = 4. All of the “interesting” hydrogn atoms could be located by analysis of potential hydrogen bonding patterns. Eltrombopag olamine Form I (Promacta®), (C2H8NO)2(C25H20N4O4) crystallizes in P21/n with a = 17.65884(13), b = 7.55980(2), c = 22.02908(16) Å, β = 105.8749(4)̊, V = 2828.665(11) Å3, and Z = 4. The initial structure solution reversed the orientation of one of the cations. Levocetirizine hydrochloride Form I (Zyzal), C21H27ClN2O3Cl, apparently crystallizes in P21/n (even though it is a chiral molecule and exhibits weak second-harmonic generation) with a = 24.1318(21), b = 7.07606(9), c = 13.5205(7), β = 97.9803(4)̊, V = 2286.38(12) Å3, and Z = 4. Edoxaban tosylate monohydrate Form I (Lixiana®), (C24H31ClN7O4S)(C7H7O3S)(H2O), crystallizes in P21 with a = 7.55097(2), b = 7.09010(2), c = 32.08420(21) Å, β = 96.6720(3)̊, V = 1744.348(6) Å3, and Z = 2. Tezacaftor Form A (Symdeko), C26H27F3N2O6, crystallizes in C2 with a = 21.05142(2), b = 6.60851(2), c = 17.76032(5) Å, β = 95.8255(2)̊, V = 2458.027(7) Å3, and Z = 4. Pomalidomide Form I (Pomalyst), C13H11N3O4, crystallizes in P-1 with a = 7.04742(9), b = 7.89103(27), c = 11.3106(6) Å, α = 73.2499(13), β = 80.9198(9), γ = 88.5969(6)̊, V = 594.618(8) Å3, and Z = 2. Palbociclib isethionate Form B (Ibrance®), (C24H30N7O2)(C2H5O4S), crystallizes in P-1 with a = 8.71337(4), b = 9.32120(6), c = 17.73722(20) Å, α = 80.0258(5), β = 82.3581(3), γ = 76.1560(2)̊, V = 1371.284(5) Å3, and Z = 2. Osimertinib mesylate Form B (Tagrisso), (C28H34N7O2)(CH3O3S) crystallizes in P-1 with a = 11.4291(3), b = 11.7223(4), c = 13.3221(4), α = 69.0246(8), β = 74.5906(7), γ = 66.4001(7)̊, V = 1511.466(13) Å3, and Z = 2. Other new structures may be discussed as they become available.

Structure determination of organic materials directly from powder X-ray diffraction (XRD) data [1,2] is now carried out extensively by researchers in both academia and industry. Most research in this field uses the direct-space strategy for structure solution [3,4] followed by Rietveld refinement. Although the structure determination process is generally carried out solely using powder XRD data, significant advantages may be gained by augmenting the process of structure determination from powder XRD data by utilizing information obtained from other experimental and computational techniques. Such multi-technique approaches are particularly advantageous in tackling complex and challenging structure determination problems, both by providing independent information that may be used directly to facilitate the structure determination process and by allowing robust validation of the final structure obtained in the Rietveld refinement. The lecture will focus on the use of solid-state NMR spectroscopy and periodic DFT-D calculations to augment the process of structure determination of organic materials from powder XRD data [5-11]. The lecture will present several case studies from recent research, including several examples of polymorphic systems and pharmaceutical materials. Recent examples exploiting the complementary advantages of 3D electron diffraction data and powder XRD data within the structure determination process will also be presented.

[1] Harris, K. D. M., Tremayne, M. & Kariuki, B. M. (2001) Angew. Chemie Int. Ed. 40, 1626.

[2] Harris, K. D. M. (2012) Top. Curr. Chem. 315, 133.

[3] Harris, K. D. M., Tremayne, M., Lightfoot, P. & Bruce, P. G. (1994) J. Am. Chem. Soc. 116, 3543.

[4] Kariuki, B. M., Serrano-González, H., Johnston, R. L. & Harris, K. D. M. (1997) Chem. Phys. Lett. 280, 189.

[5] Dudenko, D. V., Williams, P. A., Hughes, C. E., Antzutkin, O. N., Velaga, S. P., Brown, S. P. & Harris, K. D. M. (2013) J. Phys. Chem. C 117, 12258.

[6] Williams, P. A., Hughes, C. E. & Harris, K. D. M. (2015) Angew. Chemie Int. Ed. 54, 3973.

[7] Watts, A. E., Maruyoshi, K., Hughes, C. E., Brown, S. P. & Harris, K. D. M. (2016) Cryst. Growth Des. 16, 1798.

[8] Hughes, C. E., Reddy, G. N. M., Masiero, S., Brown, S. P., Williams, P. A. & Harris, K. D. M. (2017) Chem. Sci. 8, 3971.

[9] Hughes, C. E., Boughdiri, I., Bouakkaz, C., Williams, P. A. & Harris, K. D. M. (2018) Cryst. Growth Des. 18, 42.

[10] Al Rahal, O., Hughes, C. E., Williams, P. A., Logsdail, A. J., Diskin-Posner, Y. & Harris, K. D. M. (2019) Angew. Chemie Int. Ed. 58, 18788.

[11] Al Rahal, O., Williams, P. A., Hughes, C. E., Kariuki, B. M. & Harris, K. D. M. (2021) Cryst. Growth Des. 21, 2498.

Crystal engineering has emerged as an important field of solid-state chemistry, developing tools to deliberately design functional organic solids. A particularly exciting aspect of crystal engineering is the tuneability of physicochemical properties of organic solids such as solubility, thermal stability, bioavailability etc. without altering the underlying molecular structure(s) – a concept of high relevance for pharmaceutical industry.^[1] Altering physisochemical properties can be achieved by relying on different solid forms, such as polymorphs, cocrystals, and salts.^[2] The latter two are multicomponent systems that, in organic solids, are essentially distinguished by the position of a proton within the crystal structure. While different chemical systems can appear in different forms, proton transfer has rarely been observed for multicomponent systems with identical stoichiometric composition.^[3]

In this contribution, we present an extremely rare case of polymorphism between a metastable molecular (cocrystal) and ionic (salt) form of a two-component system based on nicotinamide and a dicarboxylic acid, induced by supramolecular tautomerism. In specific, we show the polymorphic transition from a metastable cocrystal to a salt, monitored using time-resolved powder X-ray diffraction (PXRD) and solid-state nuclear magnetic resonance spectroscopy. Both formerly unknown structures of were solved ab initio from PXRD data and further analyzed using spectroscopic methods, as well as density functional theory calculations.

Figure 1: Monitoring the polymorphic transition from metastable cocrystal to salt using (a) PXRD and (b) ¹⁵N ssNMR spectroscopy.

[1] Almarsson, O. & Zaworotko, M, J. (2004). Chem. Commun., 17, 1889-1896. [2] Aitipamula, S. et al. (2014), Cryst. Growth Des., 12, 2147−2152 [3] Bernasconi, D., Bordignon, S., Rossi, F., Priola, E., Nervi, C., Gobetto, R., Voinovich, D., Hasa, D., Tuan Duong, N., Nishiyama, Y., Chierotti, M. R., (2020). Cryst. Growth Des., 20, 906-915.

In the pharmaceutical crystal, hydration/dehydration phase transitions are often observed phenomena during manufacturing or storage. They lead the substantial crystal structure change, so they are critical for the important physicochemical properties that depend on the crystal structure, such as stability, solubility, and bioavailability. However, after dehydration, single-crystal integrity tends to degrade, resulting in powdery crystals. We have successfully revealed solid-state structural rearrangements using ab initio Structure Determination from Powder X-ray Diffraction data (SDPD) technique [1-4]. Interestingly, some crystals show "isomorphic desolvation," in which the XRD pattern does not change significantly after dehydration, meaning the initial molecular arrangement is well preserved. We can reveal an isomorphic desolvation mechanism by comparing the crystal structures from powdery crystals in the hydration/dehydration phase transitions, which can be achieved using the SDPS technique.

Carbazochrome sodium sulfonate trihydrate, a hemostatic agent, undergoes stepwise dehydration by humidity control or heating. The hydration number decreased from 3 to 2.5, 2, 1, and anhydrous form I under dry condition, and it showed isomorphic desolvation (Fig. 1). Their crystal structures were analysed by SDPD technique to show the API molecules are linked through Na cations to form polymeric structure, and the molecular arrangements were very similar. It is noteworthy that the first dehydration did not occur at non-coordinated crystalline water C, but water molecules A and B which coordinated to Na cation were dehydrated in sequence. This removal order was explained by the crystal structures' stability after each dehydration, calculated using CASTEP quantum mechanics calculations. Even after dehydration, the molecular arrangements were almost kept by adjusting the molecular positions slightly. After removing crystalline water C, the crystallinity degraded significantly, indicating the molecule C is essential for stabilizing the whole crystal structure. Thus, the mechanism of the stepwise dehydration behaviour, and isomorphic desolvation were revealed by SDPD technique.

Figure 1. Chemical diagram and molecular arrangement of trihydrate, dihydrate, and monohydrate phases of Carbazochrome sodium sulfonate.

[1] Fujii, K., et al. (2010) J. Phys. Chem. C 114, 580.

[2] Fujii, K., Uekusa H., Itoda N., Yonemochi E. & Terada K. (2012) Cryst. Growth Des. 12, 6165.

[3] Fujii, K., Aoki, M. & Uekusa, H. (2013) Cryst. Growth Des. 13, 2060.

[4] Putra, O. D., Yonemochi, E., Pettersen A. & Uekusa H. (2020) CrystEngComm 22, 7272.

Keywords: Structure Determination from Powder X-ray Diffraction data; dehydration; crystal structure, pharmaceutics; quantum mechanics

Part of this work was supported by JSPS KAKENHI Grant Number JP18H04504 and 20H04661 (HU).

Trichlormethiazide is a thiazide derivative, an important group of diuretic drugs, which is used in the treatment of hypertension. The Cambridge Structural Database (CSD) contains only one report (KIKCUD) associated with this pharmaceutical [1], corresponding to the orthorhombic form of anhydrous S-Trichlormethiazide. The PDF-4/Organics database contains two entries. One is the calculated pattern of the CSD entry (PDF 02-094-5865) and the other is an experimental unindexed pattern (PDF 00-039-1828). In this contribution the structure of racemic Trichlormethiazide was determined from laboratory X-ray powder diffraction data. This material was also characterized by FT-IR, TGA and DSC. The structure was determined with DASH [2] and refined by the Rietveld method with TOPAS-Academic [3]. The final unit-cell parameters are a = 8.4389(6), b = 8.8929(7), c = 9.7293(8) Å, α = 91.315(3)°, β = 106.113(2)°, γ = 97.1580(17)°, V = 694.73(9) ų, Z = 2. The refinement converged to R_p = 0.0512, R_wp = 0.0694, and GoF = 2.704. In the crystal structure, the molecules form chains along the a-axis connected by cyclic N-H···N and N-H···Cl hydrogen bonds. The chains are connected by additional cyclic N-H···Cl hydrogen bonds to form layers almost parallel to the ab plane. The fingerprint plots and energy frameworks diagrams of S and racemic forms clearly show the different intermolecular interactions and their topologies. A detailed discussion will be present in this work.

Ab initio structure determination from Powder X-ray Diffraction (PXRD) data is continuously demonstrating its merit thanks to advances in modern phasing algorithms, computing power and X-ray instrumentation. As the technique has long passed the question “can it be done”, there is another question to answer: “how far can it be pushed”. In this respect, this presentation has two aims: (i) to show methodologies that allow for solving of difficult structures of organic molecules and (ii) to highlight the level of accuracy that can be obtained from PXRD data.

This presentation is focused on structure determination of crystals that feature large molecules, disorder, or radiation-induced changes. The first part of the presentation outlines a phasing methodology that can result in an interpretable structural model. The methodology relies on the phasing process in the charge-flipping program [2] Superflip [3] by introducing a partial or incorrect structure obtained by a direct-space algorithm FOX [4]. The second part of the presentation will address structure completion and refinement, and highlight examples of how high quality data can be used for restraint-free Rietveld refinement, modeling disorder from difference Fourier map, and for obtaining insights in bond order disambiguation.

While the majority of shown examples rely on PXRD data collected at synchrotron sources, the potential of data collected in a laboratory diffractometer will also be discussed.

[1] Šišak Jung, D. et al. (2014). J. Appl. Cryst. 47, 1569-74
[2] Oszlanyi, G., Sütő, A. (2004). Acta Cryst. A60, 134-141
[3] Palatinus, L., Chapuis, G. (2007). J. Appl. Cryst. 40, 786-790
[4] Favre-Nicolin, V., Černý, R. (2004). Z. Kristallogr. 219(12), 847-856.