Prof. Detlev Grützmacher, Director of the Peter Grünberg Institute: New laser for computer chips
The Peter Grünberg Institute (PGI) at the Forschungszentrum Jülich is the central platform for basic research in the area of nanoelectronics, combining research, technological development and innovative device concepts. Dr. rer. nat. Detlev Grützmacher is acting director at Jülich's Peter Grünberg Institute.
1. Together with your partners, you presented the first semiconductor laser which consists solely of elements in the carbon group (silicon, germanium, tin) (main group IV). What are the benefits of this combination?
Prof. Detlev Grützmacher: It is only laser modules made from semiconductors in main group IV—which include silicon (Si), germanium (Ge) and tin (Sn)—that can be integrated into the silicon chip production process without any difficulties. Consequently, the compound creates a new basis for transmitting data on to computer chips via light. On the one hand this is faster and also requires only a fraction of the energy compared to the conventional copper lines. In this way we could speed up communication between logic and memory elements, which is regarded as a bottleneck for advances in computer technology.
2. What process is used to “marry” germanium and tin and how high is the tin concentration?
Prof. Detlev Grützmacher: The monocrystalline thin layers of Si, Ge and Sn are deposited directly on to silicon substrate with a special epitaxy process. We developed this special process, which we call reactive gas-source epitaxy, and implemented it on a commercially available system; we have also applied for a patent for the process.
The special feature is that we cause hydride radicals (GeH3) and chlorine compounds (SnClx) on the surface to create an exothermal chain reaction in which a high amount of energy is released. This causes the molecules on the surface to decompose and the atoms to align accurately in the crystal lattice. The substrate itself remains relatively cold; the germanium and tin atoms are effectively frozen into the crystal. This enables us to achieve tin concentrations of 12–14 percent at present, well above the solubility equilibrium, which is about 1 percent.
3. What is the difference to the lasers that are currently used in fiber optic cables?
Prof. Detlev Grützmacher: Typically, lasers that have been used in some cases for decades in telecommunication networks and data centers consist of elements in main groups III and V, we also call them III/V lasers. For optical transmission into fiber optic cables, lasers are used as individual modules that are placed on a heat sink and are not integrated into a silicon chip. Because of their different crystal properties, these III/V lasers cannot be applied directly to silicon with the required quality; instead, for the integration, they have to be produced externally at considerable expense, and then be stuck on to the wafer.
Since the thermal expansion coefficients differ considerably from silicon, these elements have a very limited service life. The material costs for these compounds, such as gallium arsenide, are usually disproportionally high. As opposed to this, GeSn can be deposited directly on Si. Another benefit: GeSn lasers emit at a wavelength that is not absorbed by the silicon and germanium electronic components on the chip.
4. What wavelengths does the laser emit?
Prof. Detlev Grützmacher: Characterization of the germanium-tin laser in Dr. Hans Sigg's working group at the Paul Scherrer Institute (PSI) in Switzerland showed that the emitted light is in the wavelength range of 2.25–2.3 micrometers. This wavelength is very dependent on the tin concentration of the GeSn alloy. For the highest tin concentrations of 14 percent that we can achieve at present we expect wavelengths above three micrometers, in other words, in a range of near to mid infrared. Therefore, you could not see the beam with the naked eye. In the future, the wavelength range could be extended and be optimized for specific applications by adding Si to the alloy.
5. What are the biggest barriers for the germanium-tin (GeSn) laser on its path from the laboratory to application?
Prof. Detlev Grützmacher: In the past, we pumped the laser optically; in other words, we generated the excited electrons needed for the laser by radiating the GeSn with a strong light source. For the application, this excitation has to be generated electrically; that is, by applying an electric voltage. Additional thin layers on the sides of the GeSn layer are needed for this. In the past, the layers had to be cooled to -180 degrees to get a laser. To achieve operation at room temperature, the tin concentration has to be increased. Together with our international partners, we are well on our way to overcoming these barriers.
6. Apart from data communication on computer chips, are other applications conceivable?
Prof. Detlev Grützmacher: In the wavelength range I mentioned, many carbon compounds also have strong absorption lines: greenhouse gases or carbonates and biomolecules, for example. Therefore, the germanium-tin compound is also suitable for innovative sensors that are located directly on the chip of mobile devices and with which these substances can be detected.