Computational Materials Physics
Staff
Profs. Graeme Ackland, Simon Bates, Mike Cates and Jason Crain; Drs. Jamie Cole, Derek Hepburn, Davide Marenduzzo, Paul Tulip and Miriam Marques
About Us
Simulation is one of the major theoretical tools we use for the understanding of complex physical systems. The group's research activities include the application of advanced computational methods to the study of molecular materials in both the solid and liquid state. Methods used include electronic structure methods such as density functional theory (DFT) and classical and ab initio molecular dynamics simulations. Additionally, there has in recent years been a shift away from purely equilibrium problems (calculating partition functions or Schrödinger energy levels) to working on systems that are out of equilibrium. Examples include soft matter systems such as colloids that are subjected to shearing; here exotic phenomena such as jamming and rheological chaos are seen. Likewise a fluid of mixed contents shows complex phenomena when subjected to a steady chemical potential gradient - this arises in drying, dissolution and mixing processes. In quantum-mechanical simulations of solid state phases the nonequilibrium dynamics of defects, such as grain boundaries, plays an increasing role, and first-principles molecular dynamics of molecules at surfaces addresses the kinetics of adsorption, desorption, and catalysis.
We have unusually strong links with in-house experimental programmes both in Soft Matter Physics and in Extreme Conditions Physics. We take every opportunity to confront our theories with the latest experimental data, and this makes for a very stimulating working environment for all concerned. Much of our work is funded, jointly with experiment, by a two-million pound ‘Programme Grant’ from the Engineering and Physical Sciences Research Council (EPSRC). We are also participants in EPSRCs ‘RealityGrid’ e-Science Testbed Project (part of a national initiative in Grid Computing) and several other national projects. Note that the University of Edinburgh has led the UK in computational physics since that field was invented, and currently hosts the National eScience Centre (NeSC).
Research Interests
Below are some examples of the kinds of physics problem now being addressed by PhD students and others within the group:
Flow of Colloids and Fluid Mixtures
Colloids in suspension can exhibit new phenomena such as jamming and arrest, either spontaneously or driven by stress. These are being studied by adapting glass transition theories from statistical mechanics. We also have simulation work on the topic, mainly using a powerful algorithm called LB (Lattice Boltzmann). In future we will address the dynamics of colloidal particles suspended in fluid mixtures. These are capable of new forms of arrest, and should form viscoelastic solids at very low volume fraction. Figure 1 shows a binary fluid mixture undergoing demixing. The interface between fluids is shown blue on one side, yellow on the other; the fluids themselves are transparent.
Blue Phases of Liquid Crystals
The ‘blue phases’ (BPs) of cholesteric liquid crystals are a spectacular example of soft material, made up by a self-sustained network of disclinations (one-dimensional topological defects, see Figures 2 & 3 for the networks of BPI and BPII, which are the cubic BPs stable in a field). At high temperature, a cholesteric liquid crystal is in the ‘isotropic phase’, with molecules pointing all over the place. Upon cooling down the sample, the molecules orient and their average direction (the ‘director field’) rotates in a helical fashion about a well-defined axis - this is the so-called ‘cholesteric phase’. Very close to the transition, a simple helical order is frustrated and it is more advantageous for the director field to rotate in a helical fashion about any axis perpendicular to a line - this complicated pattern was named a ‘double twist cylinder’. Mathematically, it is impossible to patch up different double twist cylinders without creating defects in between, and this gives rise to the disclination network observed in blue phases, and responsible for many of their remarkable physical properties.
Our current work on blue phases focuses on their domain growth and on their behaviour in a field. Very recently, we have performed large scale simulations of quenches from the cholesteric phase which lead to the formation of slowly evolving amorphous disclination networks: this could provide a model for the ‘blue fog’, of blue phase III, whose structure is to date not well understood theoretically.
Defects and Nanostructure in Materials
Modern materials design depends on knowing how higher level structure (grain boundaries, twinning, dislocations) emerges from atomic interactions. This requires both first principles quantum mechanics and molecular dynamics to span the length scales. Recent work has involved shape-memory metals which have one structure at high temperature (or B-field) and another at low T or B. Applications include motorless flapping wings for micro-aircraft and one-off unfolding of sails for space probes. The physics underlying these effects is found at the level of nanostructure; figure 4 shows a defect pattern in Zirconium. The 'bottom up' approach of predicting behaviour is gradually being replaced by top-down rational design: what atoms should we use to give the right material properties?
For more information, on this and the latest developments in related areas, see: http://www.homepages.ed.ac.uk/graeme/group.html
Molecular Physics
We are interested in understanding the relationship that exists between molecular (and electronic) structure and function. A complete elucidation and understanding of this relationship would be of great importance for molecular biology and medicine biotechnology, and other molecular-based technological devices. A particularly fruitful strategy is to examine small, relatively "simple" model systems, chosen such that, whilst being computationally tractable, they offer a sufficiently rich and diverse palette of physical phenomena that they allow an understanding of bigger, more complex systems. For example, small peptides can be viewed as fragments of larger proteins. We have a long-standing programme of research investigating such fragments using both classical atomistic and first principles quantum mechanical molecular dynamics methods, in collaboration with experimentalists at Rutherford Appleton Laboratory (RAL), and the National Physical Laboratory (NPL). esearch on such systems has provided valuable insights into the physical mechanisms by which peptides aggregate, whilst also clarifying how currently used classical potentials may be improved in order to better describe these systems.
A second, complementary strategy is to examine the behaviour of molecular crystals. Organic molecular crystals are capable of displaying a diverse range of physical phenomena, such as superconductivity, metal-insulator transitions, charge and spin ordering, Peierls transitions, etc. They thus provide an environment in which such physical phenomena may be probed theoretically and experimentally. From a technological perspective, such rich phase diagrams permit a crystal's properties to be "tuned", which may be used in, for example, sensing applications or molecular electronics. Density functional theory provides a powerful and flexible technique for the examination of such problems. Recent studies by the group have investigated pressure-induced metallisation in single-component molecular crystals, and modulating electronic properties through crystal polymorphism.
In addition to this applied research, the group has interests in developing improved simulation methods. In particular, research has concentrated upon the development and implementation of advanced sampling techniques for use in molecular dynamics simulations, new methods of extracting experimental observables from simulation data, and the development of new polarisable potential models for an improved description of liquid water in biomolecular simulations.
Copyright © 2011 School of Physics, University of Edinburgh
This page is maintained by Richard Blythe.