By Peter Weiss
From Indianapolis, at the March meeting of the American Physical Society
Microchips work properly only if the silicon they’re made of is precisely spiced with other atoms known as dopants. However, today’s chip-making methods can’t guarantee consistent doping for regions of silicon smaller than 70 nanometers on a side, says Dongmin Chen of the Rowland Institute for Science in Cambridge, Mass. That could spell trouble for the semiconductor industry, which expects to make chip components of those dimensions just 4 years from now.
Chen and his colleagues at Rowland and the Chinese Academy of Sciences in Beijing, however, report a step toward finer dopant control. By using an atom-deposition technique called molecular-beam epitaxy and by exploiting a tendency of indium atoms to bunch together, the researchers covered an entire chip with triangular arrays of exactly six indium atoms per patch of silicon 3 nm on a side. Each patch contains only 49 silicon atoms.
Still, what the team has done is not actual doping. To make that claim, the researchers will have to overlay silicon onto the indium-peppered surface, notes team member Jian-Long Li of Rowland. Moreover, other types of dopants such as antimony are also needed for making chips, but the team has yet to deposit antimony with as much finesse as the researchers have shown with indium.
