Quantum dot laser
A semiconductor laser whose gain comes from self-assembled nanoscale islands that confine carriers in all three dimensions, giving an atom-like density of states. The practical consequences: temperature-stable threshold, tolerance of optical feedback and of crystal defects, and natural operation as a multi-wavelength comb.
A quantum well confines carriers in one dimension and leaves them free in the other two; a quantum dot confines them in all three. The gain medium is a plane of self-assembled islands, typically InAs on GaAs, a few tens of nanometers across and a few high, each holding discrete electron and hole states. The density of states collapses from the quantum well's staircase toward a set of delta functions, which is the property from which everything else about the device follows.
The advertised consequences arrived roughly as predicted. Threshold current can be very low, since few states need inverting, and its temperature dependence is weak: the characteristic temperature of good quantum dot lasers substantially exceeds quantum-well values, which matters wherever the laser must live somewhere hot and uncooled. The linewidth enhancement factor is small, because the symmetric, atom-like gain spectrum decouples gain from index; a small alpha makes the laser markedly harder to drag into coherence collapse, and quantum dot lasers have been operated stably with feedback levels that would destabilize a quantum-well DFB, opening the possibility of isolator-free links. The dots also localize carriers away from crystal defects, so the devices age gracefully on lattice-mismatched substrates; this defect tolerance is why the leading epitaxial lasers grown directly on silicon are quantum dot lasers, a result the heterogeneous integration entry places among the competing ways of putting gain on silicon.
The honest costs sit in the same physics. Self-assembly produces a distribution of dot sizes, and the resulting inhomogeneous broadening spreads the gain over tens of nanometers at reduced peak value, so quantum dot lasers yield less gain per unit length than wells and saturate at lower powers. What looks like a defect becomes the second commercial application: a single-section quantum dot laser mode-locks readily across its broad gain spectrum and emits a frequency comb of tens of lines with unusually low per-line intensity noise. Such comb lasers are sold commercially and are one of the two serious candidates, against Kerr microcombs, for feeding every wavelength of a co-packaged optical engine from one external laser source.
References: Arakawa & Sakaki, Appl. Phys. Lett. 40, 939 (1982), the original proposal; Norman, Jung, Wan & Bowers, "Perspective: The future of quantum dot photonic integrated circuits," APL Photonics 3, 030901 (2018). The comb-source comparison for co-packaged optics is in Light sources for co-packaged optics.