X-ray powder diffraction and electron single crystal diffraction, although having very different methodologies in their cores, target the same material, and can deliver complimentary information for the structure characterization.

X-ray powder diffraction is a well-established technique; its performance can be exemplified by a number of impressive highlights [1-3]. A structure analysis with powder X-ray diffraction runs thorough three main steps – (i) indexing of the powder profile, (ii) structure solution, and (ii) structure refinement. The first step represents the bottleneck for the whole procedure, being associated with the inherent problem of the powder method – projection of all reflections onto a single axis. The most difficult cases represent polyphasic samples, large unit cell volumes, and low symmetry structures.

Electron diffraction method, being able to address nanocrystals individually, allows to collect 3D single crystal data from crystals with the size down to tens of nanometres [4]. A 3D reconstruction of the reciprocal space immediately delivers information on the unit cell metric. The inherent problems of electron diffraction appear at later stages, when quantification of reflection intensities is required. The strong interaction of electrons with matter gives rise to multiple scattering, which modifies intensities of reflections in a complex manner. Recently, methods for dynamical structure refinement became available [5]; still the multiple scattering contribution cannot be accounted for during the structure solution (model building) step.

In this light, an obvious beneficial combination of two techniques is the transfer of unit cell parameters, determined from electron diffraction to powder X-ray data for subsequent structure solution and refinement. This workflow will be demonstrated by examples. Beyond this combination, analysis of diffuse scattering by the two methods will be presented, and combined analysis of total scattering for PDF calculation will be discussed.

[1] Vella-Zarb, L., Baisch, U., Dinnebier, R. E. (2013). J. Pharm. Sci., 102, 674. [2] Schlesinger, C., Bolte, M. and Schmidt, M. U. (2019). Z. Kristallogr. 234, 257. [3] Spiliopoulou, M. Karavassili, F. Triandafillidis, D.-P. Valmas, A. Fili, S. Kosinas, C. Barlos, K. Barlos, K. K. Morin, M. Reinle-Schmitt, M. L. Gozzo F. and Margiolaki, I. (2021). Acta Cryst. A77. [4] Gemmi, M., Mugnaioli, E., Gorelik, T.E., Kolb, U., Palatinus, L., Boullay, P., Hovmöller, S., Abrahams, J.P. (2019). ACS Cent. Sci. 5, 1315. [5] Palatinus, L. Brázda, P. Jelínek, M. Hrdá J., Steciuk, G. Klementová M. (2019). Acta Cryst., B75, 512.

Unit cell determination, phase identification, structure determination, structure refinement. At one point of time, X-ray powder diffraction (XRPD) was the way to go for structure characterization of microcrystalline powders, despite the analyses sometimes being slow and tedious. For a long time, we have known that electron diffraction (ED) data from microcrystals are useful for unit cell and structure determination. We would still resolve to XRPD for structure refinement, because the data are kinematical and therefore simpler to model.
Over the last 15 years, developments in ED methodology, both hardware and software, have reached a point where high-quality data can be collected routinely on a large number of crystals [1, 2]. When of sufficient quality, structures refined against these data challenge the accuracy of what can be obtained from XRPD data. By combining data from different crystals using cluster analyses, we showed that even physically meaningful anisotropic ADPs can be obtained from ED data [3]. These are notoriously difficult to obtain from XRPD data.
What can we not do with ED? Through serial crystallography experiments, we saw that it is possible to collect ED data from hundreds or thousands of crystals automatically [2]. This opens the door for automated quantitative phase analysis using ED data [3, 4, 5], challenging the bulk information that can be obtained from XRPD data. Then what do we still need XRPD data for?

[1] M.O. Cichocka, J. Ångström, B. Wang, X. Zou, S. Smeets, J. Appl. Cryst. 51(6), 1652-1661
[2] B. Wang, X. Zou, S. Smeets, IUCrJ 6(5), 854-867
[3] S. Smeets, S. I. Zones, D. Xie, L. Palatinus, J. C. Pascual, S.-J. Hwang, J. E. Schmidt, L. B. McCusker, Angew. Chem. Int. Ed. 58(37), 13080-13086
[4] S. Smeets, J. Ångström, C. O. A. Olsson, Steel Res. Int. 90(1), 1800300
[5] Y. Luo, B. Wang, S. Smeets, J. Sun, W. Yang, and X. Zou, Manuscript in preparation.

Magadiite, Na₂Si₁₄O₂₈(OH)₂·nH₂O, is known as a mineral discovered at the lake Magadi in Kenya by Hans Eugster in 1967 [1]. Since then, magadiite-type materials have also frequently been synthesized in the lab and have come into focus for various applications [2-4], like CO₂ adsorbents, drug carriers or catalysts and maintain a rising interest.

Despite many attempts, the unique magadiite structure remained unsolved. Finally, a material-specific strategy based on 3D electron diffraction successfully deciphered the atomic structure [5]. In order to enable the ab initio structure solution of the electron beam sensitive material, a sodium-free dehydrated form of magadiite was synthetically isolated and, from that, it was subsequently possible to derive a structure model for the sodium form of magadiite, later successfully refined against powder X-ray diffraction data. Furthermore, a geometry optimization, simulations of spectroscopic data and calculation of charge transfer between the water molecules and the silicate layer with DFT methods confirmed the obtained crystal structure of sodium magadiite.

The structure of the silicate layer is quite complex, as it contains 4-, 5-, 6-, 7-, and 8-rings of three- and four-interconnected [SiO_4/2] tetrahedra. Seven symmetrically independent Si atoms and 15 independent oxygen sites are present forming a dense layer of considerable thickness (11.5 Å). The symmetry can be described by the layer group c211. Each layer is chiral, but the chirality of the stacked silicate layers in the average structure (F2dd) is alternated, due to the glide plane perpendicular to the stacking axis. Bands of interconnected [Na(H₂O)_6/1.5]⁺ octahedra are intercalated between neighbouring silicate layers to compensate the charge of the layers.

The detailed knowledge now achieved on the previously unknown silicate layer and the development of an adapted synthesis combined with an ammonia-based titration will have a huge impact on the research of hybrid organic−inorganic nanocomposites based on magadiite, related layered silicates and zeolite-like structures in order to design new and more efficient materials.

Figure 1. Number of publications mentioning magadiite. Inset illustrates the structure of sodium magadiite with view along [110].

[1] Eugster, H. P. (1967). Science. 157, 1177–1180.

[2] Ge, M., Tang, W., Du, M., Liang, G., Hu, G. & Jahangir Alam, S. M. (2019). European Journal of Pharmaceutical Sciences. 130, 44–53.

[3] Paz, G. L., Munsignatti, E. C. O. & Pastore, H. O. (2016). Journal of Molecular Catalysis A: Chemical. 422, 43–50.

[4] Vieira, R. B., Moura, P. A. S., Vilarrasa-García, E., Azevedo, D. C. S. & Pastore, H. O. (2018). Journal of CO₂ Utilization. 23, 29–41.

[5] Krysiak, Y., Maslyk, M., Silva, B. N. N., Plana-Ruiz, S., Moura, H. M., Munsignatti, E. O., Vaiss, V. S., Kolb, U., Tremel, W., Palatinus, L., Leitão, A. A., Marler, B. & Pastore, H. O. (2021). Chemistry of Materials [accepted].

This research was supported by the Czech Science Foundation (project number 19-08032S).

Polymorphism is a common aspect of most commercially relevant drugs. One-third of crystalline organic molecules and about half of marketed active pharmaceutical ingredients (APIs) are known to form polymorphs [1, 2]. The characterization of all polymorphic species and the understanding of the overall polymorphic energy landscape represents a prominent aspect of drug development and is crucial to establish efficacy, formulation and shelf life. Moreover, the discovery of new polymorphs with different chemical and physical properties may result in treatments that are more effective and with reduced side effects [3].

Here, we report the crystallization, structure determination and dissolution behaviour of the δ-polymorph of the non-steroidal anti-inflammatory drug indomethacin (IMC), a poorly studied polymorph first mentioned almost 50 years ago [4] and whose structure has remained hitherto unknown. δ-IMC shows a significantly enhanced dissolution rate compared with the more common and thoroughly studied α- and γ-polymorphs, potentially connected with an increased bioavailability.

Pure δ-IMC was obtained via desolvation of the methanol solvate form. Its crystallisation results in fibrous crystals that are too tiny for conventional single-crystal X-ray diffraction (XRD). Structure determination was therefore obtained on the basis of continuous three-dimensional electron diffraction (3D ED) [5], recorded by a single-electron detector [6]. The structural model obtained from 3D ED was refined using the Rietveld method against powder XRD data, following the protocol used for other pharmaceutical compounds [7, 8] and allowing the accurate determination of free torsion angles and intermolecular bonding.

The structure solution provides a solid clarification of δ-IMC spectroscopic IR and Raman data and a tentative interpretation for still unsolved indomethacin metastable polymorphs. Moreover, it explains the observed solid-solid transition from the δ-polymorph to the α-polymorph, which is likely driven by similarities in molecular conformation.

The applied procedure for structure determination may be implemented as a standard protocol for the R&D department of a pharmaceutical company.

[1] Hilfiker, R. (2006). Polymorphism: In the Pharmaceutical Industry. Weinheim: Wiley.

[2] Cruz-Cabeza, A. J., Reutzel-Edens, S. M. & Bernstein, J. (2015). Chem. Soc. Rev. 44, 8619.

[3] Gao, L., Liu, G., Ma, J., Wang, X., Zhou, L. & Li, X. (2012). Controlled Release 160, 418.

[4] Borka, L. (1974). Acta Pharm. Suec. 11, 295.

[5] Gemmi, M., Mugnaioli, E., Gorelik, T. E., Kolb, U., Palatinus, L., Boullay, P., Hovmöller, S. & Abrahams, J. P. (2019). ACS Cent. Sci. 5, 1315.

[6] Nederlof, I., Van Genderen, E., Li, Y. W. & Abrahams, J. P. (2013). Acta Cryst. D69, 1223.

[7] Andrusenko, I., Hamilton, V., Mugnaioli, E., Lanza, A, Hall, C., Potticary, J., Hall, S. R. & Gemmi, M. (2019). Angew. Chem. Int. Ed. 131, 11035.

[8] Andrusenko, I., Potticary, J., Hall, S. R. & Gemmi, M. (2020). Acta Cryst. B76, 1036.

We thank Diamond Light Source Synchrotron Facility for obtaining simultaneous synchrotron powder XRD and differential scanning calorimetry (DSC) data.

The crystal structure of a novel high pressure, high temperature tin oxynitride phase (Sn₂N₂O) was solved via Automated Electron Diffraction Tomography (ADT) [1]. The new phase was synthesized from a Sn-N-O precursor at 20 GPa and 1200-1500°C. Due to strong overlaps of symmetrically non-equivalent reflections, attempts to solve the unknown structure based on X-ray powder diffraction data were not successful. The use of the ADT method allows to collect three dimensional electron diffraction data (3D ED) from single nanocrystals in the TEM via a tilt movement of the crystal and sequential diffraction pattern acquisition [2]. Subsequently, the reciprocal space is reconstructed and unit cell parameters as well as space group information can be derived. The electron diffraction intensities can be extracted and used to solve the crystal structure via approaches like “direct methods”.

The new oxynitride phase crystallizes in space group Pbcn with the unit cell parameters: a=7.83 Å, b=5.53 Å, c=5.54 Å. The crystal structure could be solved ab initio with direct methods and refined taking both the kinematic and dynamic theory of scattering into account. It resembles a Rh₂S₃ type structure where the Sn atoms are sixfold coordinated by O and N atoms. The refined structure compares very well with DFT calculations demonstrating the quality of data achievable with ADT and its applicability for the structure solution of high pressure and high temperature materials.

[1]Bhat S., Wiehl L., Haseen S., Kroll P., Glazyrin K., Gollé-Leidreiter P., Kolb U., Farla R., Tseng J., Ionescu E., Katsura T. & Riedel R. (2020). Chem. Eur. J. 26, 2187-2194

[2]Kolb U., Krysiak Y. & Plana-Ruiz S. (2019). Acta Cryst. B 75, (4) 463-474.

3D electron diffraction (3D ED) has recently emerged as an alternative to x-ray diffraction to elucidate the atomic structure of nano-sized beam sensitive crystals¹. LD-EDT² is a recently developed low dose 3D ED technique for ab initio structure determination of beam sensitive crystals such as hydrated minerals or MOFs. Low dose conditions are achieved by optimizing exposure during specimen tilting. High quality diffraction data can be obtained from very small crystals without damaging the structure, and a precise sampling of the reciprocal space is assured by beam precession. We recently applied LD-EDT on Bulachite³, a hydrated Al arsenate mineral, to solve its atomic structure. Difficulties related to the small size of crystals as well as beam sensitivity due to the presence of H2O molecules inside the lattice were overcome by LD-EDT, where synchrotron x-rays previously failed. The resulting structure⁴ is comprised of layers containing edge-sharing Al-O octahedra, inter-connected with As-O tetrahedra by corner sharing. The localization of light atoms in the lattice showcases the potential of electron crystallography for yielding high quality diffraction data even under low dose conditions.