UK scientists have been granted £2.5 million to invent a nanomachine that can build materials molecule by molecule. 

Such a robot doesn't - and may never - exist, though it has been imagined for over half a century. But this autumn, researchers across the UK are starting work towards it, following the funding of three research projects by the Engineering and physical sciences research council. 

One of the projects, led by Rasmita Raval at the University of Liverpool, imagines creating a machine that can be instructed by computer to move molecules or atomic clusters as desired. Scanning tunnelling microscopes are already able to nudge atoms over surfaces, and image the result. But the goal now is to move into three dimensions, and to build a structural network of atoms. 

As collaborator Philip Moriarty, a physicist at the University of Nottingham, explained to Chemistry World, the idea is to blend scanning probe microscopy with self-assembly. In theory, the tunnelling electrons streaming around an incoming probe tip would alter the conformation of molecules attached to a nanoparticle.That conformational change would, like a wave of falling dominoes, force chemical or physical reactions either side of the nanoparticle. 

While Moriarty and other physicists work on refining probe tips and tuning the energy transfer from tip to nanoparticle, chemists on the team are excited by trying to organise the molecular self-assembly that a series of probes could set off. 'If it works, it will redefine nanotechnology as it should have been,' said Lee Cronin, an inorganic chemist at the University of Glasgow - referring to concepts promoted in the 1980s by US engineer Eric Drexler, who suggested that nanotechnology would create tiny machines dubbed 'assemblers' that could drag atoms and molecules around to make copies of themselves, or other useful devices. 

In a related project led by Harris Makatsoris at Brunel University, computer scientists and chemists are hoping to develop a computer language that could instruct the putative nano-assembler to work without human intervention. 

Back to biology 

A biological slant on the problem is taken by another EPSRC project, led by Andrew Turberfield at the University of Oxford. His team are copying nature's matter compiler: the ribosome, which assembles proteins from strands of messenger RNA. Turberfield told Chemistry World that the plan was to create a machine acting as an artificial ribosome. Like nature's ribosomes, it would run on an instruction tape: not RNA, but a strand of synthetic DNA, created by commercial solid-phase synthesis. The machine would read the tape, creating a strand of molecules and then linking them together in sequence, much as ribosomes do. Every component of the molecule-making factory would be a molecule itself. 

'Our ambition is to develop the system to the point where it could be distributed to end users as chemicals in plastic vials,' Turberfield proposed. The device would be a proof of concept, he said, that might be used for high-throughput production of molecules in a research lab, not for mass production. Nadrian Seeman, a chemist at New York University, said his team was working on similar concepts, though he noted from experience that 'getting chemistry more complicated than DNA chemistry to work is hard.' 

The ambitious projects, which are funded to 2010, were hammered out in January at a hotel outside Southampton, where scientists and EPSRC members took part in an 'Ideas Factory' workshop, giving them free rein to discuss bold ideas which might not have been funded under the usual peer-reviewed grant calls. 'The perception is that responsive mode funding can be risk-averse,' said Philippa Hemmings, who works in EPRSC's research and innovation directorate. EPSRC are hoping to promote 'transformative research', she explained, which would encourage scientists to think beyond their immediate research horizons and look for new collaborations (see p13). 

Richard Van Noorden