The University of Alberta and Quantum Silicon Inc. are making strides toward atomic electronics. According to the University of Alberta, atomic patterns that control electrons may reveal the potential of the “greenest electronics,” reaching applications such as neural networks for machine learning.
“Atoms are a bit like chairs that electrons sit on,” says Robert Wolkow, physics professor and principal investigator on the project. “Much as we can affect conversations at a dinner party by controlling the grouping of chairs and assigned seating, controlling the placement of single atoms and electrons can affect conversations among electronics.”
Structures under atomic control have been seen before; however, until now it was not possible to make custom patterns to develop electronic devices, notes Wolkow.
According to the University of Alberta, two key limitations impeded practical electronic applications. “The atoms would only remain in place at cryogenic temperature and could only readily be achieved on metal surfaces that were not technologically useful.”
In a proof-of-concept device, the research team created a system that was half atomic machine, half electronic circuit, and overcame its two major design hurdles. The device may also be scaled up, since it can be patterned on silicon surfaces.
“This is the icing on a cake we’ve been cooking for about 20 years,” says Wolkow. “We perfected silicon-atom patterning recently, then we got machine learning to take over, relieving long suffering scientists. Now, we have freed electrons to follow their nature—they can’t leave the yard we created, but they can run around freely and play with the other electrons there. The positions the electrons arrive at, amazingly, are the results of useful computations.”
The team has started on larger machines that will simulate neural networks. Conventional systems consists of transistors directed by computer software, however, the atomic-based machine will “spontaneously display the relative energetic stability of its bit patterns.” As a result, the neural network can be trained “more rapidly and accurately.”
The full details of the research can be found in the article, “Initiating and Monitoring the Evolution of Single Electrons Within Atom-Defined Structures,” published in Physical Review Letters.
