Research chemists have long dreamed of developing an artificial photosynthesis system that, like green plants, could capture and store the energy of sunlight in energy-rich molecules.

It's a formidable challenge, one that Deakin University chemist Professor Richard Russell compares to that of designing and building an entire city, when you are still learning how to make and lay bricks.

Plants have had at least 2 billion years to develop and refine the elaborate molecular machinery of photosynthesis, through the trial and error process of natural selection.

Artificial photosynthesis, along with challenges such as converting atmospheric nitrogen into nitrogenous fertiliser at room temperature, or creating data storage systems based on something like nature's own DNA molecule, is taking modern synthetic chemistry into a new realm of complexity, dominated by macromolecules - huge, very complex molecules - that must be assembled in huge, repetitive arrays.

"Organic and organometallic chemistry is moving into an exciting new era of large molecules that self-assemble into loosely bound arrays," Professor Russell said. "We're on the verge of understanding macromolecules in living organisms, and mimicking their functions.

The research group is well on the way to developing a molecular Lego kit: a set of simple, flexible molecules from which chemists will be able to assemble much larger and more complex molecules.

Several years ago Professor Russell and his colleagues became interested in artificial photosynthesis, but wanted to focus on an area appropriate to their skills and modest resources.

"Chemists are beginning to understand the molecular architecture of photosynthetic systems, particularly the primary energy-gathering systems that harvest solar energy.

"It was clear to us that if you're going to trap energy directly from the sun, you need to build very large light-harvesting assemblies whose components are loosely connected in space, via flexible bonds."

"To make it work, you really need a box of Lego blocks - children know you can buy a Lego kit as large as their parent's pockets will allow, but you can can get a basic kit that allows you to make quite complex structures from a relatively limited number of pieces.

"We've been trying to define what sort of pieces chemists will need to make a light-harvesting macromolecular array.

The challenge for the Deakin team was to devise a universal mechanism for linking these basic pieces together - the equivalent of the male-female "snap" system used in Lego component.

They came up with an idea for a molecular glue - a small molecule with two "sticky ends" that can be used to link curves, rods and angles in almost any conceivable combination.

"We recently found two reactions that are very broad-ranging in their potential - we simply take any two components, add our glue, and heat them to join them up. It turns out these our process is very special, because we can produce highly ordered assemblies in a controlled way. If things went together in a random way, it just wouldn't work."

The measure of the Deakin's team's achievement is that their paper announcing the advance has been among the top 10 cited papers in the field since it was published.

Since discovering its all-purpose molecule glue, the team has been using it to construct a range of complex chemical molecules. "We can now make molecules that are approximately rod-like, we know how to make curves with a controlled radius, and we know reasonably well how to make fixed angles, although they're a little harder to control," he said.

"But it doesn't really take us anywhere, unless we can give the blocks useful functions. So for the past 12 months, we've been trying to build in some really important chemistry, by adding effector components to the blocks.

"The blocks are fascinating in their own right, and we can assemble them into quite complicated molecules fairly rapidly. The more complicated the array, generally the more difficult it is to handle - chemists often have to take long, very circuitous routes in synthesising complex molecules, by alternately hiding and exposing reactive elements.

Professor Russell's team is collaborating with a research group at the University of Central Queensland, in Rockhampton, led by Professor Ron Warrener, who is trying to build molecular machines.

"It's an interesting challenge. Making a molecular machine is one thing - the problem is getting energy into it to make it do useful work," Professor Russell said.

The first successful attempt involved making a molecular switch with a "flip-flop" action like a conventional light switch. "We've been working on a very large E-shaped molecule, in which the central bar can be induced to flip.

"We have to be able to discriminate between one side and the other - the 'off' position and the 'on' position. We have done it by attaching a different effector to each rigid arm. The hinged bar switches from one arm to the other in response to a small change in pH."

Useful molecular machines, the smallest mechanical devices that could possibly be conceived and built, are still a long way from being practical, but Professor Russell's team is also working on something of more immediate practical use, using "Lego" chemistry.

The Deakin researchers have developed considerable expertise in chemiluminescence - developing molecules that fluoresce when energised. They are working on simple, cheap techniques for monitoring chemical reactions during the industrial manufacture of fine chemicals.

"Industry needs a simple probe that can be stuck in a vat and give a readout. We think we know how to do it, in principle.

"We need to coat a cheap probe with a metal complex containing ruthenium, by inducing it to complex with a polymer. When we dip the probe into the solution, it will emit light - we measure the light, and calculate the yield."