The primary focus of work in my group is on the construction, development and application of relatively simple potential models. The use of such models, in which potentially complex intermolecular interactions are expressed in terms of relatively simple energy functions, allows a wide range of systems, with potentially unique properties, to be accessed. In all cases direct links have been established with on-going experimental investigations (neutron diffraction, X-ray diffraction, electron microscopy, Raman spectroscopy, NMR... at high temperatures and pressures).
Research in progress includes:
1) The formation of low-dimensional crystal structures.
Many materials have now been observed experimentally to fill carbon nanotubes to form internal (low-dimensional) crystalline structures. We have developed models to directly observe the filling, allowing for the relative effects of the underlying thermodynamics and kinetics (i.e. the underlying physics of the confined environment) to be understood, as well as the atomistic formation mechanisms (information which is not directly obtainable from experimental studies). In the MX stoichiometry, for example, the crystallites formed fall into general classes of inorganic nanotube (INT) whose structures may be understood in terms of the folding of two dimensional sheets of varying morphology. These INTs may have unique mechanical and electronic properties which may be tunable (for example as a function of the INT morphology). Work is on-going to understand the internal crystal structure in alternative stoichiometries as well as for more complex mixtures. In addition the results from the atomistic models can be mapped onto analytic models to allow for the development of mesoscopic models to simulate nanotube bundles.
2) Network-forming liquids and glasses.
Systems such as SiO2, GeO2, ZnCl2, GeSe2 and BeCl2 can be considered as constructed from a percolating network of MX4 local coordination tetrahedral. At the extrema SiO2 can be considered as a (three-dimensional) network of strongly-bonded corner-sharing tetrahedra whilst BeCl2 can be thought of as an ensemble of weakly bonded (pseudo-one-dimensional) edge-sharing units. Systems such as ZnCl2 and GeSe2 can be considered as having intermediate structures in that they contain both corner- and edge-sharing units, the presence of which controls the network physical properties. In addition, systems of this type often exhibit structure on more than one length-scale, often on an intermediate range (i.e. on a length-scale longer than the short-range order imposed by the usual cation-anion Coulombic ordering). Furthermore, the short-range order has recently been shown (by high resolution neutron scattering experiments) to percolate to extended range (over 5nm). In our simplified models the fraction and edge- and corner-sharing units can be readily controlled and hence the effect of these units on the underlying structural and dynamic properties can be understood. The length-scales accessible allow direct observation of the extended-range order and allow for direct links to liquid state theories.
3) Extension of known phase diagrams.
Advances in experimental techniques are allowing more extreme conditions to be systematically probed. High temperatures and pressures allow previously unobserved crystal structures to be accessed. Alternatively, low density (clathrate) structures can be obtained and, under applied pressure, may lead to alternative (metastable) high density crystals. Simulation models allow these observations to be directly interpreted and, in addition, allow the atomistic phase transition mechanisms to be identified.
Systems such as Si, Y/Al/O mixtures and H2O show polyamorphism, that is, the existence of amorphous phases with identical compositions but different densities. Low- and high-density (LDA and HDA) states may be accessed directly from simple simulation models and used to help interpret experimental observation. Alternatively, signatures of polyamorphism may be detected in the liquid state (just above the system melting point) and the atomic configurations responsible for the underlying polyamorphism identified.
5) Nanoparticle self-assembly.
Solutions of nanoparticles of materials such as PbSe, CdS and CdSe are found to self-assemble to form complex mesoscopic structures (chains, belts, rings...) the details of which depend on the reaction conditions (solvent, temperature, organic adducts...). We are working to develop mesoscopic models to allow this behaviour to be modelled directly. The mesoscopic models are parameterised by reference to smaller-scale atomistic calculations creating a hierarchy of models operating on multiple length-scales.