Germanium, a semiconductor with far greater light conduction possibilities than silicon, is a step closer to providing a foundation for the light-based computers of tomorrow, thanks to a European team working to make the material laser-compatible. Today’s electricity-dependent silicon-based computer chips – with their on-going need for ever-smaller and more densely packed transistors to meet processing power demands – are rapidly approaching the limits of Moore’s law in terms of performance and speed. “The way to go in future is light,” said Richard Geiger, a doctoral student at Paul Scherrer Institute’s (PSI) Laboratory for Micro- and Nanotechnology and ETH Zurich’s Institute for Quantum Electronics. And because germanium is used already in the semiconductor industry to produce silicon chips, it is seen as a potential solution, once the material can be induced to lase – and to do so efficiently. The team, which includes researchers from Politecnico di Milano in Italy and Geiger’s colleagues from ETH Zurich and PSI, accomplished its feat by altering germanium’s optical properties through high tensile strain. The investigation is detailed in Nature Photonics (doi: 10.1038/nphoton.2013.67). Slight tension is generated in germanium when it evaporates on silicon. The researchers strengthened this effect using microbridges – germanium strips attached to silicon but which remain connected to each other via an extremely narrow “bridge.” They showed that the strain at the bridge could be intensified enough to create photons and become a direct bandgap. Light-emitting bridges of germanium can be used for communication between microprocessors, an international team of researchers from ETH Zurich, Paul Scherrer Institute and Politecnico di Milano has found. Unlike some other methods, such as one developed at MIT, where researchers used germanium with a 0.25 percent strain and relied on highly doping the material to increase gain, the approach taken by Geiger’s group started with germanium with an even lower strain, he said. “Actually, we had only 0.14 percent of strain in our germanium layers,” Geiger told Photonics Spectra. “But, with our method, we increase this prestrain by more than a factor of 22 to a uniaxial strain of 3.1 percent, which decreases the difference between direct and indirect bandgap from 136 meV to only 47 meV [and] leads to a much higher occupation of the direct bandgap compared with unstrained germanium.” The 3.1 percent strain is the maximum the researchers have achieved so far, Geiger said, adding that the result had more to do with the quality of the material than with the method itself. “If we had material with less defects, we could further increase the strain in the bridges,” he said. Applying higher strain would make the laser increasingly more efficient, which would also increase the lasing wavelength. Geiger said germanium will become a true direct bandgap material at 4.7 percent strain, with a corresponding wavelength of 3 µm that can be tuned from 2 to 3 µm and theoretically even beyond, as long as the material can withstand the high stress. And the approach is not limited to germanium on silicon, he said. “It is actually a very general technique in the way that it can be applied to any layer which is prestrained on its substrate, and where the substrate can be removed selectively. He noted that the technique can be used to alter the emission wavelength of traditional III-V laser systems by varying the strain. Now that the investigators have proved that the technique works, the next step is to make a laser by incorporating the gain material into an optical cavity. They currently are working to design such a cavity that is compatible with their bridge structures and that won’t reduce the strain levels needed for gain. “As soon as we can put all those things together at once, we will achieve the laser,” Geiger said, “which is definitely not years of research away from us. But there are still some small steps to take on our steady way towards the goal.”