The simplest quantum light source in solid-state physics is also one of the most demanding. Take a single quantum dot, excite it to create exactly one exciton, and wait. When the exciton recombines, it emits exactly one photon. No more, no less. That photon carries information about the exciton’s spin and valley state, travels at the speed of light, and can be routed through fibers and waveguides. In principle, excitons give us deterministic single photons on demand.

In practice, “single photon source” is a term of art with three separate quality metrics — purity, indistinguishability, and brightness — and a source that excels at one may fail at another. The history of exciton-based single-photon sources is the history of learning to satisfy all three simultaneously, in materials that range from cryogenic gallium arsenide to room-temperature two-dimensional semiconductors and color centers. This article explains what each metric means, why they trade off, and where the field stands on the hardest goal of all: a good source that works at room temperature.

Purity: one photon, not zero or two

Purity asks the most basic question: when the source fires, does it emit exactly one photon? The standard measure is the second-order correlation function at zero time delay, denoted g(2)(0). For an ideal single-photon source, g(2)(0) equals zero — two photons never arrive together, because there is never a second photon to coincide with the first. For a classical laser, g(2)(0) equals one. Values below 0.5 are generally taken as evidence of single-photon emission, and values approaching zero indicate high purity.

Achieving low g(2)(0) with excitons requires isolating a single emitter. In a quantum dot, this is done by confining one exciton in a nanoscale island and exciting it in a way that creates at most one electron-hole pair per pulse. Resonant excitation — driving the exciton transition directly with a laser tuned to its energy — gives the cleanest results, because it avoids creating extra carriers that could produce additional photons. Above-band excitation, which creates carriers in the surrounding crystal that then relax into the dot, is simpler experimentally but more prone to multi-photon events and timing jitter.

Background light is the enemy of purity. Any other emitter in the excitation spot, any scattered laser light, or any secondary recombination pathway contributes extra photons that raise g(2)(0). Modern experiments therefore combine spatial isolation, spectral filtering, and resonant excitation to suppress background. In well-optimized quantum-dot sources under cryogenic conditions, reported g(2)(0) values are very low — well below 0.1 and in careful experiments approaching the few-percent level or better — though the exact number depends on excitation conditions, filtering, and whether background counts have been subtracted. Readers should treat any single purity number as conditional on those details.

Indistinguishability: photons that interfere

Purity alone is not enough for quantum information. Most photonic quantum protocols — linear-optical quantum computing, boson sampling, quantum repeaters, and many sensing schemes — require that photons from different emission events, or from different sources, be indistinguishable: identical in frequency, polarization, spatial mode, and temporal shape, so that they interfere when brought together on a beamsplitter. The standard test is Hong-Ou-Mandel interference, in which two photons entering opposite ports of a beamsplitter bunch together and exit the same port, suppressing coincidence counts. The depth of the Hong-Ou-Mandel dip quantifies indistinguishability.

Indistinguishability is where excitons struggle. The exciton’s environment — phonons, charge fluctuations, nuclear spins, and spectral diffusion — perturbs its energy from one emission event to the next, shifting the photon’s frequency and broadening its line. Even if each photon is individually pure, photons emitted at different times may differ enough to degrade interference. This is sometimes described as dephasing: the exciton loses phase coherence before it recombines, and the emitted photon inherits that imperfection.

Two strategies mitigate dephasing. The first is to make emission faster than dephasing, using the Purcell effect: placing the emitter in an optical cavity or near a photonic structure that accelerates radiative decay, so the photon leaves before the environment can perturb it. The second is to quiet the environment: cryogenic cooling, clean materials, charge stabilization, and resonant excitation all reduce fluctuations. The best quantum-dot sources combine both, achieving high indistinguishability — Hong-Ou-Mandel visibilities well above 90 percent in leading cryogenic experiments — but only under carefully optimized conditions. Indistinguishability and purity must be reported together; a source with excellent purity but poor indistinguishability is of limited use for interference-based protocols.

Brightness and collection efficiency

Brightness asks how often an excitation pulse actually yields a usable photon in the desired mode — collected into a fiber or waveguide, not merely emitted in some random direction. A source can be pure and indistinguishable yet useless if most photons are lost to total internal reflection, emitted into the substrate, or absorbed before they escape. Semiconductor quantum dots sit inside high-refractive-index material, so most emitted light is trapped unless the photonic environment is engineered to extract it.

Extraction engineering is a major subfield in its own right. Micropillar cavities, photonic-crystal waveguides, circular Bragg gratings (“bullseye” structures), nanowires, and solid-immersion lenses have all been used to direct emission into a useful mode. The best devices combine Purcell enhancement with high extraction, so that a large fraction of excitation pulses produce a collected photon. Reported end-to-end efficiencies vary widely by architecture and by how efficiency is defined — source efficiency, collection efficiency, fiber-coupled efficiency — and comparisons across papers require care about which definition is used. The honest summary is that high brightness has been demonstrated in cryogenic quantum-dot devices, but it requires sophisticated nanofabrication, and it remains an active engineering challenge rather than a solved problem.

Room temperature: the hard frontier

All of the results above are easiest at cryogenic temperatures, typically a few kelvin, where phonon scattering is suppressed and spectral diffusion is reduced. The obvious question is whether exciton-based single-photon sources can work at room temperature, where they would be vastly cheaper and more practical to deploy.

The physics is not favorable, but it is not hopeless. Room temperature means thermal energy of roughly 26 millielectronvolts, which exceeds the exciton binding energy in many bulk semiconductors and guarantees rapid phonon scattering. Emitters that work at room temperature therefore need either very large binding energies, strong confinement, or both. Monolayer TMDs, with binding energies of hundreds of millielectronvolts, are one candidate. Localized defect-bound excitons in TMDs and in hexagonal boron nitride, and color centers in diamond and related materials, are others. In each case, spatial localization creates a discrete emitter whose emission can show photon antibunching — g(2)(0) below 0.5 — at room temperature.

Demonstrations of room-temperature single-photon emission from such systems have been reported by multiple groups. What has generally not yet been demonstrated at room temperature is the full combination of high purity, high indistinguishability, and high brightness in one device. Indistinguishability is the hardest to preserve: at room temperature, phonon-induced dephasing broadens the emission line far beyond the Fourier-transform limit, so photons from successive emissions differ enough to degrade Hong-Ou-Mandel interference. Cavity enhancement and Purcell shortening can help, but they must overcome a much larger dephasing rate than at cryogenic temperatures. As of this writing, room-temperature indistinguishability competitive with cryogenic quantum dots should be regarded as an active research goal, not an established capability.

Quantum dots versus defects versus 2D materials

It helps to compare the main emitter families explicitly. Epitaxial quantum dots in III-V semiconductors — gallium arsenide and indium arsenide systems — are the most mature. They offer excellent purity and indistinguishability at cryogenic temperatures, integration with semiconductor photonic circuits, and electrical control, but they require cryogenics and sophisticated growth. Colloidal quantum dots are cheaper and solution-processable, and single-photon emission has been shown, but blinking, spectral diffusion, and lower indistinguishability have limited their use in demanding quantum optics.

Defect and color-center emitters — including diamond nitrogen-vacancy and related centers, and defects in hexagonal boron nitride — offer room-temperature operation and in some cases spin-photon interfaces, but their emission properties, brightness, and reproducibility vary by defect and host. TMD-based emitters, including localized excitons in monolayer WSe2 and MoSe2, combine large binding energies, valley physics, and compatibility with van der Waals heterostructures, but they face challenges in emitter-to-emitter reproducibility, spectral stability, and integration. No single family dominates on all metrics; the choice depends on whether the application prioritizes indistinguishability, operating temperature, integrability, or cost.

What counts as a good source, and for what

The appropriate benchmark depends on the application. Quantum key distribution can tolerate moderate indistinguishability because it uses single photons as carriers of information rather than as interfering particles, but it demands high purity and high rate. Linear-optical quantum computing and boson sampling demand high indistinguishability above almost everything else, because their computational advantage rests on multiphoton interference. Quantum repeaters demand spin-photon entanglement and memory compatibility, which adds requirements beyond the photon itself. A source described as “excellent” for one application may be inadequate for another, and press coverage that omits the application context is a common source of confusion.

There is also a distinction between deterministic and heralded sources. An exciton-based quantum-dot source is deterministic in the sense that each excitation pulse can produce a photon with high probability. Spontaneous parametric down-conversion sources, by contrast, produce photon pairs probabilistically and herald one photon by detecting the other. Deterministic sources have a scaling advantage for multiphoton experiments, because the probability of generating many photons simultaneously falls exponentially with per-source inefficiency. This is why the quantum-dot community emphasizes brightness alongside purity and indistinguishability: all three multiply to determine the feasibility of large photonic experiments.

Outlook: what would change the picture

Several advances would materially change the outlook for exciton-based single-photon sources. Reproducible, spectrally stable room-temperature emitters with narrow lines would open practical deployment. Electrically driven devices — single-photon LEDs rather than optically pumped sources — would simplify integration, and electrically driven quantum-dot sources have been demonstrated at cryogenic temperatures. On-chip integration of sources, filters, routers, and detectors on a single photonic platform would move the technology from laboratory demonstrations to systems. And improved material quality in TMDs and other van der Waals materials could reduce spectral diffusion and improve emitter reproducibility.

None of these is guaranteed, but the trajectory is clear: the field has moved from demonstrating antibunching to optimizing the full set of metrics, and from isolated emitters to integrated photonic circuits. The next milestone to watch is not another room-temperature antibunching claim, but a room-temperature source that simultaneously reports high purity, measured indistinguishability, and useful brightness — with all three measured on the same device under the same conditions.

Sources and further reading

  • Review literature on single-photon sources and quantum-dot photonics, including reviews in Nature Photonics, Reviews of Modern Physics, and Nature Reviews Physics, covering purity, indistinguishability, and brightness metrics.
  • Experimental literature on Hong-Ou-Mandel interference and indistinguishability of quantum-dot single photons, in Physical Review Letters, Nature Photonics, and Nature Communications.
  • Review and experimental literature on room-temperature single-photon emitters in two-dimensional materials and hexagonal boron nitride, in Nature Materials, Nature Communications, and ACS Nano.
  • Literature on Purcell enhancement, cavity quantum electrodynamics, and photonic extraction structures (micropillars, photonic crystals, circular Bragg gratings) in Applied Physics Letters, Optica, and Nature Photonics.
  • Standard quantum optics references on the second-order correlation function g(2), photon antibunching, and Hong-Ou-Mandel interference.
  • Note on verification: quantitative values quoted above (for example, g(2)(0) thresholds and Hong-Ou-Mandel visibilities) are described as ranges and conditional on measurement details, because reported numbers depend on excitation, filtering, background subtraction, and device architecture. Any single record figure should be checked against the original paper’s methods.