The problems presented by micro-manufacturing is causing
engineers to use a whole new set of molecular tools.
Illustration by Robin Jareaux

In a world where everything from mobile phones to computers, portable music players and all manner of instruments is getting smaller, engineers need new tools to predict how materials will behave when they are being processed or shaped at scales as subtle as their molecular make-up.

Fabrication of materials, particularly metals, at the level of atoms and molecules is the basis of nanotechnology – the new generation of micro-manufacturing – and it poses new challenges for engineers who need to be able to predict how materials will behave when worked on as molecules.

Conventional cutting, grinding, hammering or welding is plain to see, but at the nanoscale manufacturing becomes invisible and the behaviour of materials and fluids when manipulated at this level is still being determined. Given their extensive history in manufacturing R&D, Swinburne University of Technology researchers are using supercomputers to develop predictive tools that will give nanoengineers and manufacturers a reliable guide on how materials and fluids will behave when worked as molecules.

"Experimental measurements on fluid-flow rates, for example, are extremely difficult at the nanoscale," says Professor Billy Todd, deputy-director of Swinburne’s Centre for Molecular Simulation. "What we are doing is seeking to understand the basis of how fluids flow when they are trapped in nanometer-sized dimensions – that is, the size of atoms and molecules themselves."

At these scales, the normal rules of fluid dynamics, which have been spectacularly successful since the 19th century, break down. "At this scale, you get very strong variations in the density of the fluid and when this happens you can no longer use these practical recipes to predict the flow properties," Professor Todd says.

The established science of fluid dynamics has a wide range of applications, including: calculating forces on aircraft fuselages; determining the optimum, safe, flow rate of petroleum through pipelines; and predicting weather patterns. Some of its principles are even used in traffic engineering, where traffic is treated as a continuous fluid.

However, nanotechnology heralds a whole new paradigm. For example, fluid dynamics can be used to demonstrate the flow of a river from a mountain to the sea. But can fluid dynamics accurately describe the movement of the individual water molecules? This would be important for biotechnologists who may need to predict the flow rate of a liquid being pumped into a carbon nanotube.

"We are trying to develop new theoretical methods where we can predict things like flow rates and stresses – different forces that are acting on the fluids, which you cannot do with conventional fluid dynamic techniques," Professor Todd says.

What the researchers are hoping to achieve is the same use of software, based on fluid-dynamics modelling, that is in common use in industries such as processing and aerospace to make technical predictions without the need for costly and time-consuming measurements. For example: ‘What is the stress on an aeroplane wing moving through the air at x kilometres an hour?’ or to predict the properties of polymers when processed into different end-uses.

"Instead of having to do the experiments, you can save enormous amounts of time, money and manpower by doing these sums on a computer," Professor Todd says. "As long as your theories are good, you can make remarkable predictions that are extremely accurate."

He believes that the emerging science of nanofluidics will become as successful in the realms of nanotechnology as fluid dynamics is already at the macro-scale.

The path to developing these tools is through a great deal of theoretical work. The researchers, all physicists and mathematicians, develop mathematical models that, in turn, will be used to develop scientific software.

Also working on the project are Dr Jesper Hansen and Professor Peter Cadusch from the centre, and a group led by Associate Professor Peter Daivis at RMIT University.

"The RMIT and Swinburne groups collaborate closely on all aspects of the project," Associate Professor Daivis says. "We meet regularly and thrash out new ideas and then vigorously discuss the results of our computations, resulting in a highly productive research collaboration."

While the group has developed many models, which they have successfully tested on known results and published their results, there is still some work to do before this translates into commercial software.

Still at the nanoscale, Professor Todd and his team are working on another project investigating the operating mechanisms and dynamics of motor proteins. These molecules, which supply cells with energy, have potential application in targeted drug delivery.

"Rather than devising an artificial motor, these proteins give us one that is naturally occurring," Professor Todd says. "There is great potential to use this naturally occurring biological protein as a motor on a nanoscale. You can use it as a source of propulsion. There is a long way to go between these fundamental studies and their application as molecular propellants, but at least the underlying mechanisms are known to some extent."

A group comprising Dr Ming Liu, head of the centre Professor Richard Sadus and Professor Todd, are collaborating with Professor Ray Norton at the Walter and Eliza Hall Institute of Medical Research to investigate a motor protein known as ATP-synthase.

Once again, the work involves developing mathematical models to explain how the motors behave. The group has developed and successfully applied a model which allows them to predict the molecule’s rate of rotation.

Using a technique known as coarse graining, that groups numbers of atoms together, the simulation is greatly simplified.

"If you wanted to simulate the molecule on a computer it is very difficult because of the huge number of atoms involved, but if you group, say, 100 atoms together as one blob and the next 100 atoms as another blob, it is simpler to model two blobs than 200 atoms," Professor Todd says. "This is the general idea of coarse graining – you reduce the complexity while still maintaining the essential physics.

"We have been able to coarse-grain certain proteins, investigate the type of motion and compare that to experimental measurements. Comparisons with experimental data are very promising."

Both of the projects are being funded by grants from the Australian Research Council.

Director of the Centre for Molecular Simulation Professor Richard Sadus says the work in both nanofluidics and molecular motors highlights the centre’s emphasis on conducting research that provides new and important insights into molecular processes.

"Knowledge of biological systems, such as ATP-synthase, provides valuable insights into nanoscale proportion required by bottom-up nanotechnology applications," Professor Sadus says. "A feature of all the centre’s work is the rigorous application of scientific principles and the use of high-performance supercomputing. The supercomputing facilities are central to the success of the research, but the critical ingredient is the talent and creativity of our research staff."