Developing structure determination from powder X-ray diffraction data for applications in molecular organic materials

The research presented in this PhD thesis is based on the theme of crystal structure determination from powder X-ray diffraction (XRD) data. More specifically, the work here advances existing methodologies primarily by implementing information from other complementary structural techniques such as solid-state NMR spectroscopy, density functional theory calculations, structure prediction calculations, and electron diffraction data. The materials studied here are a range of molecular organic solids, ranging from well-known pharmaceuticals to materials with interesting optoelectronic properties. Chapter 1 provides information on the background of the materials of interest in this study. The phenomenon of polymorphism is discussed as well as further information regarding biominerals and amino acids, whilst concluding with the aims of the research presented in this work. Chapter 2 details the experimental techniques used in the scope of this work. This includes powder X-ray diffraction, solid-state NMR spectroscopy, density functional theory calculations, electron diffraction, thermogravimetric analysis, and crystal structure prediction calculations. Chapter 3 discusses the conventional approach of using powder XRD data for crystal structure determination. This details the entire process, from peak selection and indexing all the way through to structural refinement by the Rietveld method, with subsequent structural validation using a variety of experimental and theoretical methods. Chapter 4 provides insight into the crystal structure of alloxazine and riboflavin. In the riboflavin section, the effort in determining the crystal structure is detailed, providing an alternative structure to that published by Guerain et al. with evidence from complementary techniques to discriminate between the two. In the alloxazine section, the crystal structure was determined with relative ease; however, determining the tautomeric form present within the crystal structure (alloxazine or isoalloxazine) required using other experimental and theoretical methods. Chapter 5 details the discovery and subsequent crystal structure determination of a new polymorph of L-tyrosine using powder XRD data and electron diffraction data. The crystal structure was validated using solid-state NMR data and periodic DFT-D calculations. The use of crystal structure predictions calculations was explored, leading to the discovery of multiple potential polymorphs of L-tyrosine. Chapter 6 details the elucidation of the crystal structure of racemic lysine. Several hydrates are also discovered, with their crystal structures determined. One of the hydrates is isostructural to the pure form and exhibits disorder between two arrangements; periodic DFT-D calculations were used to successfully model this disorder. iv Chapter 7 describes the preparation of a new polymorph of the popular analgesic ibuprofen. The crystal structure of this polymorph is determined from synchrotron powder XRD data and subsequently validated using periodic DFT-D calculations, with subsequent comparison to the two known polymorphs. The crystal structure is noticeably higher in energy than both two known polymorphs, determined through periodic DFT-D study. Finally, Chapter 8 describes two materials, DBTMA and PDICH, and the respective strategies used to determine their crystal structures, primarily using powder XRD data. They both show molecular similarity defined by rigid cores. Periodic DFT-D calculations were imperative in scrutinizing the molecular geometry of the DBTMA molecules, where powder XRD was inadequate. In conclusion, this thesis has described how structure determination from powder XRD data can be augmented by complementary methods to extract direct or indirect structural information. By employing methods such as 3D-ED data, it has been shown how to avoid typical bottlenecks such as indexing or structure solution, which are often impossible to overcome with imperfect data. Theoretical techniques such as periodic DFT-D calculations have been employed to probe plausible structures, and to rigorously validate or correct them, which is extremely difficult to perform with powder XRD data solely. Future work to follow on from this thesis would clearly have to focus on two aspects. One aspect being the application of the methodology presented in this thesis towards molecules of different natures, or more complex natures to those studied here. This can include inorganic or organic extended solids such as zeolites, metal-organic frameworks, or covalent organic frameworks. However, the studies of molecular organic solids with a significantly greater structural complexity e.g. peptides with ten or more amino acids, would be a huge challenge for this methodology. The other aspect would be to develop the methodology presented in this thesis even further. For example, optimizing the experimental set-up for 2D-PXRD to allow for more ideal data to be collected and used in the structure determination process. Another area for significant development would be the use of crystal structure prediction in the indexing stage, for example predicting a variety of crystal structures and using the information of their hypothetical unit cells and crystallographic symmetry to aid the indexing procedure, particularly in the cases of impure data or poorly crystalline samples