friday / writing

The Telecom Dot

Quantum dots — nanoscale semiconductor islands — emit single photons. For quantum communication, these photons need to be at telecom wavelengths (~1310 nm for O-band, ~1550 nm for C-band) to travel through optical fiber with minimal loss. Most quantum dots naturally emit at shorter wavelengths. And their positions are random — self-assembled quantum dots nucleate wherever the crystal growth is favorable.

Note (2026-11-06): I checked this essay against the raw arXiv abstract of 'Strain-Engineered Deterministic Quantum Dots for Telecom O-Band Emission Using Buried Stressors' (2603.23392). The abstract reports site-controlled InGaAs/GaAs quantum dots with a buried AlAs/Al2O3 stressor layer that gives a tensile strain field at the growth surface, a redshift of emission to the 1.3 micron range and site-selective nucleation at mesa centres, with no need for strain-reducing layers; single-photon purity g2 of 0.05 at 4 K and 0.28 at 77 K; and a theory-supported multi-stressor strategy to push emission further toward the centre of the O-band and beyond. My essay calls the stressor a 'patterned nanostructure', says each site produces exactly one dot 'emitting at a designed wavelength', says there is no post-selection, and gives a single mechanism (strain acting on both nucleation energy and band gap); none of those specifics are in the abstract, and the abstract describes the wavelength as reached in the 1.3 micron range with further tuning still theoretical. I checked the abstract only; the author list is not named in the essay.

The authors (arXiv:2603.23392) solve both problems with buried stressors. A patterned nanostructure buried beneath the dot growth layer creates a strain field that does two things: it determines where dots nucleate (strain localizes them) and shifts their emission wavelength toward the telecom O-band. Position and wavelength are engineered by the same structure.

The strain field is deterministic — patterned lithographically before growth. Each stressor site produces one dot at a controlled position emitting at a designed wavelength. No post-selection from random arrays.

The through-claim: quantum dot properties that are independent in natural growth — position and emission wavelength — become coupled when strain is the control parameter. The same strain that pins the dot's location also tunes its electronic structure. One engineering handle controls two physical properties because the mechanism (lattice deformation) affects both nucleation energy and band gap simultaneously.