Now we’ve all heard that DNA could be the future of computing, in fact Sam wrote an article for the Guardian about a recent development that showed it was possible to make DNA into components such as toothed gears and pipes.
However, it has to be said I have remained a little sceptical, I think it has something to do with the fact that, after reading the stories, I often come away with weird pictures in my head – like miniscule versions of existing machines made out of DNA.
However, last week a paper was published in Nature Nanotechnology which made more sense to me than the others. It convinced me that DNA could actually be useful in computing – to make microchips.
The paper describes how DNA can be made to “self assemble” into shapes, which are then mounted on silicon and can act as tiny fastening posts for nanoscale electrical components. Because the DNA shapes are so small, they can be placed very close together, opening up the possibility of being able to build much smaller computer components. In fact, the researchers say each component could be as little as six nanometers apart, eight times closer than is currently possible.
The technique involves taking a long strand of DNA in solution, and adding shorter ‘staple’ strands. Due to the complimentary base pair nature of DNA, the shorter strands fold the long strand of DNA into shape. The solution is then poured over a silicon substrate.
This silicon substrate requires lithographic pre-treatment to make etched ‘sticky sites’, otherwise the DNA shapes would stick randomly. However, the sites are negatively charged and would therefore repel the negatively charged DNA. The silicon is therefore washed with magnesium chloride, to allow positively charged magnesium ions to attach to the etched surface. The magnesium ions then attract the DNA shapes, which bind strongly and become mounted on the silicon.
There are still a few problems with this new technology though. Since the only shapes currently possible are triangles, or shapes composed of multiple triangles, any attached components will be able to point in any one of three directions – not ideal when the direction of the nanowires and other components need to be accurate. The researchers will also need to carry out extensive research into what the best conductor is for this application. Scientists put the timescale of the development of this technology at 10 years.