We are interested in the theory and simulation of soft materials like liquids, glasses, gels and colloids. These systems represent an important class of materials that are found throughout nature and have important applications in industry and technology. We are currently studyingnucleation phenomenain these systems to understand what controls complex structure formation and why some crystal structures are formed preferentially when others might actually be more thermodynamically favourable. This is a particularly interesting problem in the freezing of nanoparticles where there are a large number of different solid cluster structures with similar free energies. If a liquid does not nucleate to a crystal as it is cooled, it can form anamorphous solid or glass.We are studying the way particles pack into random arrangements, and how these packings are related to the properties of the glass and supercooled liquids, as a way to resolve many of the fundamental problems these interesting forms of matter pose to our understanding of thermodynamics and statistical mechanics. Another area of interest to our group is thedynamics of confined liquids.Geometric constraints on the way particles can move can have a considerable impact on their properties and we are exploring ways to control these effects.
The limit of stability of a supercooled liquid is not well understood and the existence of a spinodal singularity for crystallisation or an ideal glass transition are still open questions. However, we do know that as the liquid becomes more deeply supercooled, the unusual heterogeneous dynamics of the metastable liquid are going to heavily influence the nucleation process. We are current using simulations of silica, Lennard-Jones liquids and simple lattice gas models to study how glassy physics and nucleation are coupled.
Most recently, we have become interested in studying the freezing of nanoclusters. The most energetically stable structure of a cluster varies as a function of the number of atoms or molecules. For small clusters, the Mackay or anti-Mackay icosahedra are usually the most stable structures, with the magic numbers corresponding to completed icosahedral shells being particularly stable. As the clusters become larger, decahedra and eventually the face-centred-cubic (fcc) structures become energetically favourable. However, molecular dynamics (MD) simulations of cooled clusters show that nanoclusters generally freeze to a metastable state rather than the energetically most favored one, suggesting kinetic factors play an important role in determining nanoparticle structure. We are using computer simulation, combined with phenomenological theories, to understand the competitive nucleation between the different structures with an aim to being able to control which crystal structures are formed.
We are also studying deliquescence and heterogeneous nucleation involving soluble nanoscale particles. These small soluble particles are an important component of atmospheric aerosols and pollution but their extremely small size means that they behave differently from larger aerosol particles. We are using molecular dynamics to follow the formation of thin films on their surfaces and developing thermodynamic models to predict their unique properties for use in larger scale atmospheric models.
Glasses and Packing
If a liquid is cooled rapidly enough to avoid freezing, the molecular motions of the metastable liquid continue to slow down until the system falls out of local equilibrium to form a glass that has the mechanical properties of a solid, but the microscopic structure of a liquid. Kauzmann pointed out, in what has become known as the Kauzmann paradox, that the entropy of many supercooled liquids decreases more rapidly than the crystal so that if the glass transition did not intervene, the entropy difference between the fluid and crystal would become zero at the Kauzmann temperature, leading to a break down of the third law of thermodynamics at T=0K. Just above the glass transition, supercooled liquids exhibit a range of unusual dynamical behaviours, including non-exponential relaxation and spatially heterogeneous dynamics where some regions of a liquid move rapidly while others move slowly.
One the main goals of our research is to understand the Kauzmann paradox and its implications for the thermodynamics and dynamics of glass forming liquids. The potential energy landscape, along with the inherent structures formalism originally developed by Stillinger, offer a promising paradigm for the study of supercooled liquids. In this approach, each configuration of the liquid is represented as a point in the high-dimensional N-body potential energy function of the system that can be uniquely mapped to the closest mechanically stable packing or inherent structure. For a system with a soft potential, an inherent structure represents a local potential energy minimum while for a hard particle system, it is a collectively jammed packing and a local density maximum. The configurations that map to the same inherent structure can be grouped together into a local basin of attraction and the thermodynamics and dynamics of the liquid can be described in terms of the number of inherent structures and the motion of the system through the resulting potential energy or packing landscape of basins and saddle points. Howerver, enumerating the packings and understanding their proerties remains an unsolved problem. We are using confinement of the liquid phase as a way of obtaining some extact descriptions of the packing landscape. For example, trapping two dimensional hard discs between two narrowly separated walls has enabled us to construct the entire distribution of packings for the system by considering the local packing conditions for the particles. We are now able to explore the relationship between the dynamics of the system and how it samples its fully characterised packing landscape and to test recently introduced concepts such as the J-point.
A fundamental understanding of the way particles pack also serves as a starting point for the study of athermal systems such as sand and foams. For example, sand flows like a liquid when poured, but it can quickly become jammed in a solid-like structure. Foams can support a yield stress, but will flow like a liquid when sheared above a certain threshold. In both cases, the particle-particle structure and particle dynamics are similar to that observed in atomic and molecular liquids and glasses. This has led to the suggestion that many of these athermal systems exhibit a wide variety of physical phenomena that were originally thought to only occur in traditional thermal systems. Classical thermodynamics and thermal statistical mechanics cannot be used to explain the behaviour of a granular system because the energy required to move a macroscopic particle far exceeds that available from thermal excitation. We are using our knowledge of particle packings to develop a statistical mechanics of athermal systems which will help us understand how these materials behave in a variety of industrial settings.
Dynamics in Confined Systems
Particles, atoms and molecules that are geometrically confined in narrow channels, porous materials or between membranes, exhibit dynamic properties that are very different from their bulk counterparts. The long time limit of the mean squared displacement (MSD) of a tagged particle in a bulk fluid increases linearly with time and is governed by the well known Einstein relation. However, Brownian particles constrained to a single file, like a channel, and unable to pass, feel the permanent caging effects of their neighbours so their MSD becomes proportional to the square root of time in the long time limit. Research in our group has focused on understanding the nature of the dynamical transition that occurs as the system moves between the bulk and this confined regime. We find particles confined in channels just wide enough to allow them to pass each other exhibit different dynamics in the short, intermediate and long limits. We also show that confinement at this passing threshold can cause particles that only differ in size by a very small amount to move at drastically different speeds and we are exploring the possibility of using this as a method for separating mixtures of particles.