This Small Business Technology Transfer (STTR) Phase I research project aims at developing a practical silicon laser with the ultimate goal of monolithic integration of electronics and photonics on a single silicon substrate. Past approaches to silicon light sources, including Si/Ge superlattices, porous silicon, silicon nanocrystals, a variety of silicon-rich oxide structures, bulk silicon with a textured surface, and various optical pumping schemes have made noteworthy progress through the last few decades. Nonetheless, an electrically pumped silicon laser with satisfactory quantum efficiency has yet to be demonstrated. The proposed program exploits a recent development based on doped silicon nanostructures formed from mixtures of spin-on-dopant and spin-on-glass, which has already achieved an external quantum efficiency of 0.013% and obvious linewidth narrowing above a clear threshold, all at room temperature. The prior development, however, suffered from a poor waveguide structure which collected generated photons inefficiently and lowered effective gain. This project will attempt to solve this problem by investigating the spatial gain profile and designing a waveguide structure tailored to maximize the overlap of the optical mode field of the waveguide and the spatial gain profile. The proposed program aims at enhancing the external quantum efficiency to a few percent, which becomes comparable to compound semiconductor lasers.
Having electronics and optics work together on one silicon chip has been the vision of generations of scientists and engineers. Developing an electrically pumped silicon laser is a crucial step toward realizing this vision. Yet the intrinsically weak photon emission capability made the use of silicon problematic. A silicon laser would enable the integration of all optoelectronic components on a single silicon chip. Such chips may find applications in computers, consumer electronics, and medical devices. A special feature of the proposed silicon laser approach is its simple fabrication process, which is readily compatible with modern silicon VLSI technology. This would hasten adoption of the technology into the marketplace.
