fcurtis_physics_2017.pdf (21.02 MB)
Structure Prediction of Molecular Crystals from First Principles with the GAtor Genetic Algorithm Package
thesis
posted on 2019-11-13, 19:27 authored by Farren CurtisMolecular crystals are a versatile class of materials with applications ranging from pharmaceuticals to organic electronics. Because molecular crystals are bound by
weak dispersion interactions they often crystallize in more than one solid form, a phenomenon known as polymorphism. Understanding polymorphism has become an increasingly important issue because di fferent crystal forms may display vastly different physical properties, which aff ects their functionality for a given application. Crystal structure prediction (CSP), or the prediction of a molecule's putative crystal structures solely from its chemical composition, is a coveted computational tool as it
can predict previously unobserved polymorphs and serve as complementary tool for experimental investigations. CSP is difficult in part because one needs to sample a large con figuration space for even the simplest molecules. Furthermore, the di fferences between polymorphs can be even lower than 1 kJ/mol, making reliable CSP an extremely challenging task. In this thesis, I develop and apply a fi rst principles genetic algorithm (GA) for CSP called GAtor, which nds the most stable crystal structures for small (semi-)rigid molecules solely from their chemical composition. State-of-the-art dispersion-inclusive density functional theory (DFT) is applied for the final ranking of putative crystal structures. A preliminary version of GAtor was used to participate in the Cambridge Crystallographic Data Centre's sixth blind test of organic CSP methods. The relative stabilities and electronic properties of potential polymorphs of tricyano-1,4-dithiino[c]-isothiazole generated therein are investigated in an additional study. The methodology of the production version of GAtor, and
its corresponding initial pool generation package Genarris, are presented and applied to a chemically diverse set of four past blind test targets: 3,4-cyclobutylfuran, 5-cyano-3-hydroxythiophene, 1,3-dibromo-2-chloro-5-uorobenzene, and tricyano-1,4- dithiino[c]-isothiazole. GAtor successfully predicts the experimental crystal structure(s) for all four targets, as well as other important low-energy structures. Notably, the lowest energy putative crystal structure for 5-cyano-3-hydroxythiophene has not been reported in any previous investigations of this molecule. This may motivate
additional computational and experimental studies of this molecule.
weak dispersion interactions they often crystallize in more than one solid form, a phenomenon known as polymorphism. Understanding polymorphism has become an increasingly important issue because di fferent crystal forms may display vastly different physical properties, which aff ects their functionality for a given application. Crystal structure prediction (CSP), or the prediction of a molecule's putative crystal structures solely from its chemical composition, is a coveted computational tool as it
can predict previously unobserved polymorphs and serve as complementary tool for experimental investigations. CSP is difficult in part because one needs to sample a large con figuration space for even the simplest molecules. Furthermore, the di fferences between polymorphs can be even lower than 1 kJ/mol, making reliable CSP an extremely challenging task. In this thesis, I develop and apply a fi rst principles genetic algorithm (GA) for CSP called GAtor, which nds the most stable crystal structures for small (semi-)rigid molecules solely from their chemical composition. State-of-the-art dispersion-inclusive density functional theory (DFT) is applied for the final ranking of putative crystal structures. A preliminary version of GAtor was used to participate in the Cambridge Crystallographic Data Centre's sixth blind test of organic CSP methods. The relative stabilities and electronic properties of potential polymorphs of tricyano-1,4-dithiino[c]-isothiazole generated therein are investigated in an additional study. The methodology of the production version of GAtor, and
its corresponding initial pool generation package Genarris, are presented and applied to a chemically diverse set of four past blind test targets: 3,4-cyclobutylfuran, 5-cyano-3-hydroxythiophene, 1,3-dibromo-2-chloro-5-uorobenzene, and tricyano-1,4- dithiino[c]-isothiazole. GAtor successfully predicts the experimental crystal structure(s) for all four targets, as well as other important low-energy structures. Notably, the lowest energy putative crystal structure for 5-cyano-3-hydroxythiophene has not been reported in any previous investigations of this molecule. This may motivate
additional computational and experimental studies of this molecule.
History
Date
2018-01-05Degree Type
- Dissertation
Department
- Physics
Degree Name
- Doctor of Philosophy (PhD)