Crystal structure prediction of organic semiconductors

Abstract

This thesis presents the use of crystal structure prediction (CSP) in the evaluation and design of novel organic semiconductors. Heteroatom substitution into common organic semiconductors (pentacene in this thesis) offers a way of modulating their crystal packing and electronic properties. Initially CSP was performed on six human designed molecules and the charge mobility of their predicted crystal structures was calculated. The packing landscapes changed significantly from the unsubstituted pentacene. We found that seven nitrogen atoms led to a landscape showing a range of packing motifs, while seven nitrogen atoms favours the adoption of sheet-like motifs. Substitution patterns expected to result in the highest mobilities were found to perform worse than assumed, showing the importance of tuning both molecular electronic properties and crystal engineering. A genetic algorithm was then developed to generate new nitrogen substituted pentacenes. A population members fitness was calculated using two molecular properties important for electron transport in organic semiconductors. Five runs of the genetic algorithm gave 12 promising candidates for CSP and mobility calculations. The packing landscapes were similar to those of the seven nitrogen substituted human designed molecules. One genetic algorithm molecule showed a high number of high mobility structures close to the global minimum, making this molecule an attractive target for synthesis. Extensions to include CSP within the fitness function of the genetic algorithm represents possible future work. Addition work included the design and testing of structure generator for the generation of trial crystal structures during a CSP. The novel structure generator performed well in locating the experimental structures of three test molecules and was used in the group's submission to the 6th blind test, of which one molecule is also presented here. The experimental structure of this molecule was located in lists ranked by lattice energy and free energy, though the free energy list ranked the experimental structure as the global minimum.

More information

Identifiers

ORCID for Graeme M. Day: orcid.org/0000-0001-8396-2771

Catalogue record

Export record