Sunlight is bright, abundant and famously unruly. A laser is narrow, directional and coherent, which is why it has long been the standard pump for producing entangled photons. In a new experiment, researchers instead collected daylight, isolated a thin violet slice and used it to drive the same nonlinear process. The result was a pair state with 93.9% fidelity to the ideal Bell state they targeted.
The peer-reviewed Optica study is a striking proof of principle, but its two headline numbers require different interpretations. Fidelity describes the quality of the reconstructed quantum state. It is not the fraction of sunlight converted into photon pairs. The reported efficiency is comparable to a laser only after accounting for the very different pump bandwidths.
What the 94% result actually measures
The researchers reconstructed the two-photon state from coincidence counts recorded in 16 polarization settings. Its fidelity to their target Bell state was 0.939, with an uncertainty of 0.027. Rounded to a percentage, that is 94% ± 2.7 percentage points. They also reported concurrence of 0.905 ± 0.053 and purity of 0.919 ± 0.045.
Those measurements answer related but distinct questions. Fidelity asks how closely the measured state overlaps the intended maximally entangled state. Concurrence quantifies the degree of entanglement on a scale from zero to one. Purity indicates how much the reconstructed output resembles one clean quantum state rather than a statistical mixture. Together, they describe highly entangled, relatively low-noise photon pairs.
They do not show that 94% of incoming solar photons became useful pairs. Most sunlight never reached that point because of collection losses, filtering, polarization and the crystal’s strict phase-matching requirements. The 94% figure concerns state quality among the detected pairs, not energy conversion or yield.
How the researchers turned daylight into a quantum pump
This was not a crystal held casually in a sunbeam. The team collected direct sunlight with a one by 1.4 metre Fresnel lens covered by colour films, then used a short-pass filter, a dichroic mirror and a custom glass concentrator. A multimode fibre with a 50-micrometre core carried the selected light to the photon source.
There, filters narrowed the spectrum to 405 nanometres with a bandwidth of 1.5 nanometres. Polarization optics prepared equal horizontal and vertical components before the beam entered a periodically poled potassium titanyl phosphate crystal inside a polarization Sagnac interferometer. Through spontaneous parametric down-conversion, some 405-nanometre pump photons produced pairs near 810 nanometres. The full experimental report documents the optics and measurements taken on three separate days.
The source avoided an electrically powered pump laser, but that is not the same as an electricity-free quantum instrument. The crystal still needed temperature control to maintain phase matching. The avalanche photodiodes, time-tagging electronics, power meter and solar-tracking hardware also need support in a working system. The achievement is specifically a sunlight-pumped entangled-photon source.
For each polarization projection, the team collected data for two minutes and searched for paired detections inside a one-nanosecond coincidence window. After normalizing for the changing sunlight power, the average was about 10 coincidences per minute for every 100 nanowatts of pump power. Counts far outside the expected arrival-time peak were close to zero, indicating little accidental background.
Why incoherent light can still produce entanglement
Sunlight is spatially and temporally incoherent. At first that sounds incompatible with entanglement, but the relevant degree of freedom matters. Pump coherence in one degree of freedom limits the entanglement attainable in that same degree of freedom. Here the target entanglement was in polarization, while sunlight’s most obvious disorder was spatial and spectral-temporal.
The researchers filtered and polarized the light, then arranged two down-conversion pathways so the detectors could not tell which path produced a pair. Their superposition created the polarization-entangled state. Earlier work had already shown that a broadband free-running diode source could pump high-quality polarization-entangled pairs. The conceptual door to less coherent pumps was open before this experiment; the first claimed here is natural sunlight driving polarization-entangled SPDC.
That distinction matters. The result does not overturn the value of lasers, whose directionality, stability and tunability remain enormously useful. It shows that laser-like coherence is not a universal prerequisite for every form of photonic entanglement.
The Bell test supplied a second check
State tomography reconstructed the density matrix, but the team also tested the correlations using the Clauser-Horne-Shimony-Holt form of Bell’s inequality. Local classical descriptions are bounded by S = 2. The sunlight experiment obtained S = 2.5408 ± 0.2171, exceeding that boundary by 2.49 standard deviations.
Bell tests became central to quantum physics because they distinguish correlations permitted by local hidden-variable models from those predicted by quantum mechanics, work recognized by the 2022 Nobel Prize in Physics. Here, exceeding the classical boundary reinforces the conclusion that the measured correlations were genuinely nonclassical.
This was not a loophole-free Bell test, and 2.49 standard deviations is modest beside the strongest foundational experiments. That does not negate the result because the paper’s main goal was demonstrating a new source, not closing every Bell-test loophole. It does define the proper confidence level for reporting the result.
“Comparable efficiency” comes with an important footnote
The team measured about 1,600 coincident pairs per second per milliwatt of pump power. In the same apparatus, laser pumping had produced about 7,500 per second per milliwatt. By the raw per-power figure, sunlight was therefore lower than the laser, not equal to it.
The authors’ comparison rests on effective bandwidth. The sunlight pump occupied a band 1.5 nanometres wide, while the crystal’s useful phase-matching window was only about 0.2 to 0.3 nanometres. Much of the measured solar pump power sat at wavelengths unable to participate. A narrow laser concentrates far more of its power inside the usable window. Once the rates are normalized for that spectral mismatch, the researchers describe them as comparable.
This is also different from matching the brightest purpose-built laser sources. A 2023 laser-pumped Sagnac source, for example, reported 96.72% fidelity and was engineered for very high brightness. The sunlight experiment indicates no fundamental penalty from low pump coherence once usable spectral power is considered. It does not set an overall performance record.
What the demonstration changes, and what it does not
The work brought together researchers at the University of Ottawa, Max Planck Institute for the Science of Light, Friedrich-Alexander University Erlangen-Nürnberg and University of Rochester. The Max Planck group’s publication summary points to sustainable quantum applications in resource-limited environments.
Possible destinations include remote terrestrial stations and spacecraft that already receive abundant sunlight but place a premium on electrical power, mass and laser durability. Those uses remain proposals. The current apparatus needed direct sun, careful alignment, substantial filtering and long acquisitions. Its pump power also varied with weather and solar irradiance.
There is nevertheless a practical engineering lesson. Better concentrators, tracking, coatings, larger collection fibres and nonlinear materials with better-matched acceptance could capture more usable light. Entangled pairs already matter in fields such as quantum-enhanced microscopy, communication and sensing, although a solar source would require substantial development before joining those systems.
This is one study, not settled consensus. Its durable contribution is narrower and more interesting than saying sunlight has replaced the laser. It experimentally separates two ideas often treated as inseparable: a pump can be incoherent in space and time yet still support high-quality entanglement in polarization. The Sun supplied the photons. A carefully designed apparatus supplied the indistinguishability that made their quantum relationship possible.
