Molecular crystals offer a rich vein of physical phenomena considerably different from the optical, electronic and mechanical properties exhibited by conventional solids such as covalent or ionic crystals. This arises partly because of the generally weak intermolecular interactions such as van der Waals and dipolar interactions , and partly because of the subtle interplay that may arise between inter- and intra-molecular degrees of freedom.
In recent years, molecular crystals have become the subject of intense theoretical and experimental research [2,3,4,5,6,7,8,9,10]. Much of this has centred around developing an understanding of the physics of these systems in order to harness it for technological applications such as molecular electronics [6,11,12,13], and to this end, there have been concerted efforts to attempt to understand the nature of charge carriers in these systems. An area of equally intense activity is that of ``crystal engineering'' or super-molecular chemistry [14,15]. This fascinating area is based upon the use of individual molecules in order to attempt to design functional molecular crystals that may be used for a variety of technological applications. The key behind this philosophy is that the crystal structure and packing formed by an array of molecules is determined to a large extent by the intermolecular interactions such as hydrogen bonds and van der Waals interactions; thus one can, given enough understanding of these interactions, manipulate them advantageously in order to produce materials with technologically useful properties, such as non-linear optical response .
Molecular crystals of organic molecules are also interesting from the biological and biochemical perspective. An understanding of the hydrogen bond is essential to comprehend many important biological processes and molecules such as proteins and peptides. Molecular crystals offer an ideal test-bed in which to understand the nature of the hydrogen bond in biochemical systems in more detail.
It is possible to define a molecular crystal as a solid that is formed by electrically neutral molecules interacting via weak non-bonding interactions, primarily van der Waals. If the constituent molecules possess specific functional groups then the possibility also exists for the formation of hydrogen bonds and dipolar interactions that will also serve to stabilise the crystal. In general, there is little electronic charge overlap between molecules, and therefore the constituent molecules retain their identity to a large extent. This is in contrast to covalent or ionic solids, where the individual properties of constituent particles in the crystal are completely lost . This has led to a number of studies in which the solid state environment is treated as a perturbation to a molecular calculation , or in which ab initio or experimental molecular charge densities are used in conjunction with classically derived intermolecular potentials in order to study the behaviour of molecular crystal systems .
Approaches such as these have the advantage of being relatively simple and computationally cheaper than attempting a full ab initio treatment of the molecular solid; however, considering that the molecular structure itself may change in response to the crystal environment, in addition to the actual electronic density itself, questions must arise over the validity of such treatments. Furthermore, such approaches are based upon the idea of zero overlap between molecular wavefunctions; this implicitly assumes that space may be partitioned into Wigner-Seitz-type cells associated with each molecule . Such an approximation may not always be valid, and neglects quantum mechanical interactions such as the exchange repulsion. For these reasons, a full ab initio approach is more appealing: the crystal itself is dealt with quantum mechanically, and no assumptions need to be made about the nature of the interactions or the level of molecular overlap; rather only the cell contents need be specified. This is conceptually more satisfying, and aesthetically more appealing.
The work in this thesis concerns applying density functional theory, and density functional perturbation theory, to amino acids in both the solid and gaseous states. In particular, the emphasis is on the determination of the geometric, electronic, dielectric and vibrational properties, and elucidating the effects of the crystal structure upon the molecular properties. The amino acids are the building blocks of peptides and proteins, and, given the increasing interest in the life sciences evidenced in projects such as the Human Genome Project , are candidates for ab initio calculations. A thorough understanding of their detailed physical and chemical properties as provided by ab initio calculations will allow a complete understanding of the biological processes that they participate in.