Summary of grant
This was a JREI grant for the purchase of a multi-processor computer to carry out a series of projects in the area of molecular dynamics, quantum mechanics and chemical informatics. The science involved a mixture of high-performance computing calculations to be carried out by specialist simulators (Dr Wilson (molecular dynamics/ Monte Carlo), Prof. Hutson (quantum mechanical studies of intermolecular forces), Drs. Clark and Tozer (density functional theory)) and calculations to be carried out in conjunction with complementary experimental studies (Prof. Harris (solid state NMR), Prof. Howard (X-ray Crystallography), Dr. Fox (NMR/synthetic work), Dr. Cooper (X-ray and IR studies).
Quantum mechanical calculations have been carried out to develop force fields to describe the interactions of many common liquid crystal molecules. The force fields have been used to simulate bulk nematic phases using molecular dynamics methods and calculate some of the key physical properties that determine the behaviour of liquid crystals in displays (e.g. rotational viscosity coefficients).
A new parallel molecular dynamics program, GBMOL_DD, has been developed. The program uses a technique known as domain decomposition that greatly improves the parallel performance on a multi-processor machine. It can be used to simulate molecules with arbitrary connectivity composed of both spherical and nonspherical interaction sites. The program is ideal for the simulation of coarse-grained models for systems such as liquid crystals, oligomers, polymers, dendrimers and large polyphilic molecules; and is likely to provide a major impact in high performance computing over the next few years,
Calculations have been carried out using a new morphing procedure designed to developing accurate potential energy surfaces for complex systems. The potential surface developed for He-OCS, is already being used extensively by other groups in calculations on liquid helium droplets. In addition, quantum dynamical calculations have been carried out for a variety of systems, including large clusters containing open-shell molecules, floppy Van der Waals trimers such as Ar2HF (that are important in studies of nonadditive intermolecular forces) and fully coupled non-adiabatic calculations on Br-HF.
Quantum mechanical calculations have been carried out for systems of ultracold atoms and molecules. These include the first ever calculations of low-temperature reactive scattering in an alkali atom + diatom system; calculations of potential energy surfaces in systems such as K + K2 and Li + Li2; and calculations on interactions between molecules such as NH and Rb atoms, which are important in current attempts to achieve sympathetic cooling of such molecules by thermal contact with a laser-cooled Rb gas.
Detailed investigations have been carried out to improve the quality of exchange-correlation functionals used in density functional theory (DFT) calculations. DFT studies are the most widely used quantum mechanical calculation in many areas of chemistry, and the new functionals developed can dramatically improve the quality of these calculations. In particular, a newly developed functional, B97-2 gave chemical reaction barriers that are twice as accurate as the popular and widely used B3LYP functional. A hybrid functional (1/4 GGA) was developed also that gave improved predictions of bond lengths and vibrational frequencies. A new approach, denoted multiplicative Kohn-Sham (MKS), was developed for calculating NMR shielding constants directly from theoretical electron densities.
An extensive series of calculations have been carried out in support of experimental (NMR, X-ray and synthetic) studies within the department. These have included calculations of nuclear shieldings (and hence chemical shifts) for a wide range of organic/pharmaceutical compounds in crystalline solids for comparison with experiment. This has proved valuable in trying to understand the relation of chemical shifts to molecular geometry, particularly when different polymorphs exist. High accuracy quantum calculations coupled with calculations of proton, boron-11 and carbon-13 chemical shifts have been used to determine molecular geometries for a range of new carboranes and related species. Simulations of oriented X-ray diffraction patterns have been used to model the surface regions of polyethylene films, and help interpret experimental X-ray data from these systems.