Scheme: University Research Fellowship
Organisation: University of Bristol
Dates: Jan 2012-Sep 2015
Summary: Is glass a true solid?
We know that windows in old buildings are thicker at the bottom than at the top because glass is a liquid that flows down slowly over the centuries. At least we would know that if it were true. In fact before Pilkingtons developed plate glass flat planes could not be made and the windows have always been uneven from new. But there is a grain of truth in the story because although glass does not move on measurable timescales the true nature of glass, whether it is truly solid or a very viscous liquid is still unknown. We have taken a step towards solving this riddle and it seems that glass does become a true solid if cooled very slowly indeed.
Unlike when liquid water suddenly freezes to solid ice, a glassy liquid gradually becomes more viscous as the temperature is reduced. The picture above shows a computer simulation of a viscous liquid indicating solid-like and liquid-like regions.
The question is what happens at very low temperature, and does the whole material become truly solid? Combining computer simulation and information theory originally used for breaking the Enigma code, we aimed to predict whether a viscous liquid in fact becomes truly solid. In an interesting twist of fate we found that the size of the solid-like regions increases and that atoms in the solid-like regions organize into icosahedra which are like pentagons in three dimensions and, like pentagons, don’t tesselate.
We found that the size of the solid regions of icosahedra would grow until eventually there would be no more liquid regions and so the glass should be a true solid.
Dates: Apr 2007-Dec 2011
Summary: Local structure in liquids out of equilibrium
When a liquid is cooled below its freezing temperature, it can either freeze into a crystal, or ‘choose’ not to, and become a disordered solid, or glass. Common glass, silicon dioxide, is just one example of a generic state of matter. Unlike crystals, glasses are not in equilibrium, in general they ‘want’ to be crystalline, but are ‘frustrated’. Neither crystallisation nor glass formation are understood, although they have been observed for 4000 years. Apart from curiosity, we need to understand glass formation and freezing because:
(1) With more understanding of glass formation, we could design new materials. For example, metallic glasses promise large improvements in mechanical properties. One example of metal failure is the first jet airliner, the Comet whose tragic accidents were caused by metal fatigue. Normal metals fail at the boundaries between the microscopic crystal grains, each grain is a crystal lattice, at the boundary, lattices of different orientations form weak points. Glasses have a disordered structure, so have no grains nor grain boundaries and are less prone to failure.
(2) Understanding crystallisation is key to tackling the protein problem. According to biologists, ‘structure is function’: to find the structure the protein must be crystallised. Of the 60,000 proteins that comprise the human genome, only 20% have been crystallised. Until more proteins can be crystallised their function –and purpose – remains elusive.
We study the role of local structures of molecules in glass formation and crystallisation, using computer simulation and novel experiments based on 3D imaging of individual nanoparticles which form nano-particle crystals and glasses. Our work reveals a much higher level of detail than conventional methods, and can determine the origins of the ‘frustration’ which keeps glass-forming materials from crystallising, and how the local structure of glasses and liquids is different.