What’s Happening? – Theoretically


Every discovery is usually a product of theory and experiment, and the MacDiarmid Institute’s new deputy director, Shaun Hendy, knows only too well how important both are for science. “More and more happens in silicon rather than in the lab these days,” he says of his successful and productive collaborations with experimental physicists in a range of research areas.

As a theoretical physicist and computer modeller, Shaun Hendy has been part of the institute from its very beginnings in 2002. His joint effort with University of Canterbury physicist and MacDiarmid principal investigator Simon Brown pre-dates the institute’s founding, but has flourished under its collaborative atmosphere.

At the Christchurch laboratories, Simon Brown’s team is busy fabricating nanoparticles and constructing nanowires. Shaun Hendy’s job is to improve the process by explaining theoretically what’s actually happening in details. “Microscopy can only go so far,” he says. “There are two thrusts to my approach. One is computational, using simulations on the scale of individual atoms, and the second is the use of computer models as an overview of a complex process. That allows us to study nanoparticles in realistic conditions, but faster and cheaper.”

It means huge computer power and giga bytes of data, but the results can improve experimental design and point to new research questions. “Simon uses a bottom-up approach, where the materials assemble themselves, and with the help of computer simulations we’ve developed tricks to get them to assemble in ways you want.”

The nanowires are made by blowing tiny clusters of metallic nanoparticles into trenches and grooves etched on to an insulating substrate. If the velocity of the nanoparticles is tuned correctly, they will bounce off some parts of these pre-fabricated surfaces, and stick to others. “We did lots of simulations of that and found that the most successful process is when the particles bounce twice. They need to be at the right speed, but because it’s a sticky environment, they slow down as they bounce and it’s a bit more controlled.”

However, letting the computer run through the simulations is the easy part. The hard work is working out what the results mean and forming a theory which can then be applied to the experiment. As a result of this effort, the team is not only making the wires now, but has developed other methods of using templates such as, masks coated in non-stick polymer to stop the particles from clogging it, and surfaces coated with a polymer that repels the particles so that they only stick to polymer-free areas.

“There are still things we don’t understand. For example, one of the open questions is the effect of spin on the properties and behaviour of particles. But the more we understand the system, the better we can work with it.”

Another research focus that evolved from the collaboration is the analysis of phase transitions during melting. When nanoparticles collide, they don’t behave like larger-scale materials. “When nanoparticles collide they coalesce, a bit like droplets of water, to reduce their surface area. The collision releases enough heat to melt the particles.”

Computer simulations showed that the best explanation is the difference in the volume to surface ratio. In nanoparticles, the surface area is large in relation to the particle’s volume, and that modifies the physics. “With macroscopic materials you can almost ignore the fact that you have a surface. But with nanoparticles, almost half of all atoms could be on the surface, so the properties change.”

Shaun Hendy says an even more unusual phenomenon was that some particles remained stable beyond their melting point. “What was really curious was that although we could tell they were melting, they behaved differently from macroscopic particles. They were still solid at higher temperatures, where they’d normally be liquid. They superheated.”

Viewed from a macroscopic perspective, the melting point of a particle decreases with its size – the smaller the particle the lower the melting point, because there are proportionally more atoms in the surface and it takes less energy to break the chemical bonds between them. “But as always in science, there are counter examples – and again it’s to do with surface properties.”

Shaun Hendy found that when nanoparticles of aluminium melt, the molten part doesn’t wet the solid surface. “Instead it beads up like water on Gortex fabric. That’s quite unusual. In order to melt it has to go through the phase transition but because of this, it can remain stable beyond the melting point.”

Whether or not the nanoparticles are attached to a surface also makes a difference. If they are in contact with a surface, they can take heat from the “thermal reservoir” provided by it, and that can deliver enough energy to melt them. If the particles are free, however, they can stay in a superheated state. “Not all nanoparticles can be superheated in this way, and we’re now in the process of trying to categorise which properties make it possible.”

Although Shaun Hendy describes this aspect of his research as a curiosity and mostly of theoretical interest, he says it adds to the understanding of the processes involved. “One of the problems of working with nanoparticles is that they are very unstable because their melting point goes down the smaller they are. While it’s hard to see how this feature could be exploited immediately, it’s an important aspect to understand.”

His most recent research interest takes Shaun Hendy from nanowires to nanocapillaries. He says the obvious application driving this interest is the ongoing search for cheaper and faster ways of sequencing DNA. “People have been developing smaller devices that can handle smaller amounts of DNA. We’ve looked at the uptake of nanodroplets by carbon nanotubes and it was purely by accident that we stumbled upon this when we realised that metal droplets can be absorbed by carbon nanotubes. The simulations showed some surprising results and now we have learned some tricks to manipulate the process and make carbon nanotubes draw fl uid as well as remove fluid.”

The surprising results point back to the wetting properties of materials. “When people were fi rst making carbon nanotubes they thought it would be nice to fill them with metals. That way the carbon would protect the metal and you’d get nice thin wires. But metals don’t wet carbon, they just bead up.”

However, using computer models, Shaun Hendy was able to show that that was only the case if you dipped a carbon nanotube into a vat of molten metal. “If you take only a small droplet of metal it works. It’s the surface tension of the droplet itself that drives the metal into the nanotube capillary.”

The result of all this is another method of making nanowires, and it led to more modelling work to study the growth process of nanotubes in the presence of metals.

As the new deputy director of the MacDiarmid institute, Shaun Hendy says his focus will be on ensuring that the same good mentorship he received is available to younger researchers in the Institute. He studied mathematical physics at Massey University with Paul Callaghan as a lecturer, and later, when he joined Industrial Research Ltd after fi nishing his PhD in Alberta, Canada, he worked with Jeff Tallon. “Good mentoring was so important for my career. I want to ensure that this continues to be available for the people who’ve joined the institute as PhD students or postdocs and are now ready to become independent researchers.”

He sees a steady period ahead for the Institute, with funding secured for another six years. He doesn’t want to make any big changes, but says that he and institute director Richard Blaikie share an interest in the economic impact of the institute’s research output. “We know that science is important for economic growth but its impact is difficult to quantify. So we are trying to do that, to get a better idea of what return we get on the investment in our science. As part of a bigger picture I’m particularly interested in how science drives innovation.”