Vella-Zarb, Liana (2009)
Ph.D. thesis, University of Birmingham.
The successful prediction of the crystal structure and symmetry of a material can give valuable insight into many of its properties, as well as the feasibility of thermodynamically stable poly-morphs to exist. It is not uncommon, however, for numerous theoretical structures to be found within a narrow energy range, making absolute characterisation of the crystal structure impossi-ble. The aim of this project was to investigate a number of structures from this scenario, high-lighting the key differences between three potential methods for the automated comparison of predicted and experimental crystal structures. This work was carried out by comparing the simulated powder diffraction patterns of theoretical predicted crystal structures of small organic materials with their experimental powder diffraction patterns, so that the experimentally identified structure could be automatically singled out from the many calculated. The use of traditional agreement factors (eg. Rwp) was compared with more sophisticated approaches namely PolySNAP, which uses principal-component analysis, and Compare.x, an algorithm based on weighted cross-correlation. Five structures were analysed, two of which had not been previously characterised. As the structure prediction calculations are carried out at 0K, and experimental data were collected over a range of temperatures (10K-293K), the effect of the resulting variations in lattice parameters on the automated processes is discussed. In all cases, Rwp has proven to be a poor and unreliable discriminator in the comparison of pre-dicted and experimental structures. The more contemporary methods based on PolySNAP and Compare.x both gave encouraging results when used to study the three known structures imida-zole, chlorothalonil and 5-azauracil, and they have consequently been used in the successful so-lution of the two previously unknown structures adenine and guanine. A difference in sensitivity in the matching of data collected at different temperatures between the latter approaches was noted. It was found that although there is considerable overlap between the two methods, they are not absolutely interchangeable, and this distinction may be exploited in future work where more case-specific comparisons are carried out. Automated comparison techniques cannot yet replace visual comparison completely, but they reduce it drastically. Ultimately, comparisons made computationally serve as a complement to human judgement, but they may not yet elimi-nate it.
This unpublished thesis/dissertation is copyright of the author and/or third parties. The intellectual property rights of the author or third parties in respect of this work are as defined by The Copyright Designs and Patents Act 1988 or as modified by any successor legislation. Any use made of information contained in this thesis/dissertation must be in accordance with that legislation and must be properly acknowledged. Further distribution or reproduction in any format is prohibited without the permission of the copyright holder.
Repository Staff Only: item control page