Rethinking and redesigning the semiconductor laser/quantum noise controlled semiconductor lasers

The present canonical design of the semiconductor laser (SCL), in force since The early 1970s, is incompatible with high coherence (narrow linewidth). The reason is fundamental and is a consequence of the quantum-mandated umbilical relationship between induced emission (gain) and spontaneous emission (noise) exacerbated by the modal concentration of optical energy in the high loss III-V material. We demonstrate, theoretically and experimentally, a new design paradigm which results in over three orders of magnitude reduction in the spectral linewidth of the SCL compared to commercial lasers now deployed. The key difference is a radical (some 99% in our case) transfer of stored optical energy from the III-V material to nearly transparent silicon which is an integral part of the laser resonator. This laser should constitute a serious candidate to take over the role of the Distributed Feedback (DFB) SCL as the light source for future coherent optical networks.