What does it take to make the semiconductor laser a high coherence laser international semiconductor laser conference

The Semiconductor laser (SCL) is, arguably, the most important player in the optoelectronic field. It is hard to imagine a modern communications measurement, or a sensing system without it. It owes this distinction principally to its monolithic semiconductor character, which is responsible for a long list of crucial attributes. These include small size, efficiency, natural compatibility with electronic driving circuitry, speed, structural and chemical control of key features. Another feature of the SCL, which is mentioned less often is its intellectual elegance. Its theoretical underpinnings, design and fabrication require an intricate interweaving of solid state physics, quantum field theory, semiconductor device theory, material science, and laser theory. The chemical and fabrication control enables us to vary the active medium from that of a bulk semiconductor to that of atom-like quantum dots. The incorporation of spatial modulation, of the structure, modulated gratings, photonic crystals, for example, enables a spatial control that would be analogous, to the ability to design crystals with varying size of atoms and of periodicities. The noise, and the resulting degraded coherence, of the semiconductor laser is an example of the Dissipation-Fluctuation theorem. Which links losses with noise. This is manifested in the SCL by following chain of causally related events: high optical losses (dissipation) ⇢ large compensatory gain provided by the inverted population of electrons and holes ⇢ high rate of spontaneous recombination emission into the laser mode ⇢ low coherence. This chain can, however, be snapped by taking advantage of the new flexibility afforded us by the Si photonic platform. This is achieved by redesigning the laser mode so that the overwhelming majority (∼ 99%) of optical energy is moved away from the lossy III–V material into the, essentially lossless, Si. The residual, about 1% in our case, of the optical energy remaining in the III–V is just sufficient to provide the now reduced, threshold gain. Applying these ideas results in new lasers in which the fundamental quantum noise is some three orders of magnitude below that of high-performance commercial Distributed Feedback SCLS. Some thoughts of future directions for improved coherence in SCLs will conclude the talk.