Changing Phase

Richard Bowles first met Shaun Hendy (now the Deputy Director of the MacDiarmid Institute) when they did a summer internship together at Industrial Research Limited in 1994. The two young scientists went their separate ways, doing very different things, but met again at a conference in Los Angeles not that long ago, and realised that their interests had converged.

“It was logical that we do some work together,” says Bowles, who came over from Canada in May 2010 to work with Hendy and his group.

An Associate Professor in Theoretical Chemistry at the University of Saskatchewan, Bowles is interested in the chemical and physical properties of the molecules in liquids, particularly the processes involved when liquids change phase. He wants to know how things transform from random to ordered structures and uses a combination of statistical mechanics, classical thermodynamics, and computer simulation to understand what gives rise to the complex collective properties of “soft” materials such as glasses, colloids, gels and nanoparticles, at a molecular level. “In the type of research I’m looking at, there are a few hundred atoms of molecules involved. The type of structures formed in these very small systems are very different to those found in bulk materials, where there are millions upon millions of atoms.”

Not surprisingly he has a few projects on the boil.

The first relates to very small systems of gold. “Gold nanoparticles are made a lot, and are used for a variety of things. They have different properties,  depending on their structure,” says Bowles. If he can understand what drives the formation of these different structures, then perhaps a scheme can be implemented to direct and ultimately control those structures.

Bowles’ second interest is in atmospheric clouds, which contain a complex mixture of water droplets, pollutants and dust particles, all in a variety of  different sizes ranging from a few microns down to just a few nanometres. The phase behaviour of these components, whether the water droplets are liquid drops or solid ice crystals, and the distribution of particles’ sizes are important data used in climate models. Predicting the size distribution is one challenge, but Bowles wants to know what it is about the structures of very small particles that changes their properties and how they behave in clouds. “While the usual thermodynamic assumptions used in climate change models work very well in predicting the properties of clusters down to about five nanometres, below that they systematically get worse,” says Bowles. “Some models predict tens of these droplets per unit volume, while others suggest there are tens of thousands or hundreds of thousands, giving a very different prediction.” By understanding how small clusters form within atmospheric clouds, Bowles hopes his work will ultimately improve these climate models.

The focus of Bowles’ third line of work is creating nanoparticles which are uniform in size or structure. Essential for making the nanomachines of the future, the process of making nanoparticles usually results in a distribution of sizes. Making them uniform is a big challenge. One way to do this may be through the use of nano- or micro-fluidics. In a very wide channel, the particles hardly notice the walls and move by randomly hopping past their neighbours so their mean squared displacement, a measure of how far particles have moved on average, increases linearly with time. When confined to a very narrow channel the particle flow is much slower. Constrained to a single file, and unable to pass, they feel the permanent caging effects of their neighbours and their mean squared displacement only increases as the square root of time. By choosing the channel width just right, some of the particles feel like they are in a wide channel and move quickly, while slightly larger particles feel like they are confined and move slowly, even though the particles are in fact very similar in size. Geometric constraints on the way particles can move can have a considerable impact on their properties and by understanding the optimal geometry of these tubes, Bowles is aiming to force different sized nanoparticles to travel down a narrow channel, at different speeds, allowing the components of the mixture to be separated out by size.

Bowles is also looking at the way particles pack together in different materials, for example, the difference between structured crystals, such as diamonds, and the more random packing of glasses and amorphous structures such as those formed by silica dioxide in window glass. The packing of a material results in very different properties, and relates to thermodynamics.

In particular, he’s intrigued by the Kauzmann paradox. “There are three laws people live by in a world of thermody namics, which should be obeyed,” says Bowles. For example, the third law of thermodynamics states that the entropy of a perfectly ordered substance tends to zero as the absolute temperature approaches zero. When a liquid is cooled to its freezing temperature, it orders to form a crystal, causing its entropy to decrease suddenly. The entropy of the crystal then continues to decrease to zero as its temperature is decreased further. However, some liquids can be supercooled below their melting temperature and remain liquid. The entropy of these liquids decreases faster than that of the crystal and so could potentially become lower than that of the crystal at a temperature above zero. This would eventually cause problems with the third law since the ordered crystal should have the lowest entropy at zero degrees Kelvin, so something must happen to prevent this. Experimentally, the problem is avoided. Bowles wants to know how this paradoxical situation can be resolved.

Here in New Zealand, Bowles is looking at yet another related topic. He’s working with Hendy and his research group until about February or March 2011, and halfway into his trip, they appear to have found a problem to collaborate on, namely the melting and freezing of nanoparticles.

“Smaller sized nanoparticles should melt at a lower temperature than the bulk material, but some don’t,” say Bowles. For example, gallium and aluminium nanoparticles seem to melt at higher temperatures. One possibility is that this class of materials is non-wetting. “The liquid of this material doesn’t really like the solid of the material. They don’t form a film,” says Bowles. It also affects the way these materials freeze. “What is it about these systems that make them do this?” he asks.

Using computer models and simulations, the collaborators have some calculations in mind for this class of materials that they’d like to solve, looking at things like the nucleation barrier. Melting usually occurs in a continuous way, but these systems may have a free energy barrier to melting “We want to calculate these barriers, how these processes occur, and how it should happen,” says Bowles. “We’re working on similar problems, but using different technology and ideas. We can learn these different techniques, and choose the best one to solve the problem,” says Bowles.

While his work obviously has some associated technological outcomes, Bowles is much more focused on understanding general principles. “I’m interested in solving a problem, a concept, understanding a whole class of materials, rather than a particular substance,” he says. “These materials should all be understood by the same set of principles.”