published_or_final_version / abstract / Chemistry / Doctoral / Doctor of Philosophy

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

Computer simulations provide a powerful tool for the investigation of liquid crystalline phases. In this thesis the ability of simulations to calculate accurately the values of material parameters of liquid crystal molecules is investigated. An all-atom force field for liquid crystal molecules is developed using first principles density functional theory calculations on small organic molecules, which encompass key structural features of a range of common liquid crystalline molecules. Molecular dynamics simulations of these 'fragment' molecules are carried out in the liquid phase to test the force field parameters by determining densities and heats of vapourisation. Good agreement is found between experimental values and those calculated from simulation. Equilibrium molecular dynamics (MD) calculations were then performed for the nematogen n-4-(trans-n-pentylcyclohexyl)benzonitrile (PCH5). These simulations were performed using a fully atomistic model for several temperatures. The MD trajectories were used to obtain densities, order parameters, and values for the rotational viscosity coefficient 71. Several methods of obtaining 71 were tested based on the director angular velocity correlation function, the director mean-squared displacement, and statistical mechanics methods based on the rotational diffusion model. Good agreement is obtained between experimental values of 71 and those found from simulation. Further MD simulations of PCH5 using a 216 molecule system and the force field derived in this thesis were carried out to calculate the flexoelectric coefficients e(_s) and e(_b) for PCH5. The temperature dependence of and was examined along with the separate contributions to e(_s)and e(_b)arising bom the electrostatic and van der Waals interactions. The calculated values of e(_s)and e(_b) are consistent with available experimental data. The van der Waals and electrostatic contributions are found to be of similar magnitude and opposite sign.

- Alicyclic diols can hydrogen bond in many different ways and yield most interesting structures. In this thesis, eight C2-symmetric diols 48-50, 78, 79 and 81-83 were synthesized and their crystal structures were determined. No less than seven of these show unusual solid state behaviour: 48 and 78 are inclusion hosts; 49, 50 and 78 form doubly-stranded hydrogen-bonded ladder structures, where there is a strong preference for each strand to be homochiral; 78, 81 and 82 undergo self-resolution during recrystallization; and 83 forms chirally pure crystals (but the material is still racemic). - One of the favourable supramolecular synthons for hydroxy compounds is the (O-H)6 cycle of hydrogen bonds. When this cycle is formed by a racemic compound, its enantiomers alternate down-up-down etc. around the cycle. No case of an (O-H)6 cycle involving chirally pure hydroxy compounds is known. These observations indicate a strong preference for the (O-H)6 cycle being constructed from achiral or racemic molecules rather than from chirally pure hydroxyl compounds. Racemic (??)-48 and (??)-92 which are already known to form (O-H)6 cycles in the solid state were prepared in chirally pure form and their X-ray crystal structures determined. No (O-H)6 cycles were observed for these homochiral diols. These findings confirm that the (O-H)6 motif occurs only for achiral or racemic compounds. - Similarly, the edge-to-edge eight-membered aryl C-H???N dimer involves either achiral molecules or those of opposite chirality. No chirally pure dimers of this type are reported. Racemic compounds 42-44 that are known to pack using the C-H???N dimer were synthesized in chirally pure form. No edge-to-edge eight-membered aryl C-H???N dimers were formed in the solid state. Hence this supramolecular synthon is only favoured for achiral or racemic compounds only. - Other major conclusions are that the cause of self-resolution is due to packing energy. In some cases it is likely that solvent choice, or solvent plus temperature selection, can be used to control self-resolution.

Thesis (Ph.D.)--University of Delaware, 2007. / Principal faculty advisor: Suresh G. Advani, Dept. of Mechanical Engineering. Includes bibliographical references.