Quantum technology has long assumed that entanglement, the connection between particles that underpins quantum computing and secure communication, requires a laser. Lasers are coherent: their waves march in step, single-colored, and that orderliness has seemed necessary to coax nonlinear crystals into splitting photons into entangled pairs. A team at the University of Ottawa and the Max Planck Institute for the Science of Light has now broken that assumption in the most direct way possible, by pointing a solar concentrator at the sky and using plain sunlight to generate entangled photons outdoors.
The experiment, published in Optica, culminates a decade-long argument within Robert Boyd’s group that coherence is not actually the requirement. Entanglement can live in one property of light, polarization, even when the light is completely disordered in other properties, like color and direction. The team first showed this with LEDs, which are incoherent but produce polarization-entangled photons. Sunlight was the harder test: it spreads in every direction and contains a rainbow of colors, and some of the field’s most prominent researchers doubted whether any photons at all, let alone entangled ones, would emerge from a sunlight-driven nonlinear process.
The setup is an exercise in concentrating chaos. A Fresnel lens roughly the size of a household window gathers direct sunlight; a cone-shaped glass concentrator, built by Hanieh Fattahi’s team at the Max Planck Institute, funnels the light into an optical fiber about as wide as a human hair. The concentrated sunlight, still highly incoherent in space and time, is polarized and directed into a periodically poled potassium titanyl phosphate crystal, where spontaneous parametric down-conversion splits individual pump photons into pairs of photons entangled in polarization.
The results clear the bar for genuine quantum entanglement. Quantum state tomography reconstructed the two-photon state with a concurrence of 0.905 plus or minus 0.053, on a scale where 1 is maximal entanglement, and a fidelity of 0.939 to the ideal Bell state. The measured CHSH Bell parameter, 2.54, exceeds the classical limit of 2 by about two and a half standard deviations, ruling out any local hidden-variable explanation. The pair generation rate, normalized to pump power, is comparable to laser-pumped sources.
The significance is partly practical, partly conceptual. Lasers draw watts of electricity to deliver milliwatts of optical power, and the cryogenic and vacuum infrastructure of quantum systems adds further energy costs; the information and communication technology sector already accounts for a few percent of global greenhouse gas emissions. A sunlight-pumped source is the first entangled photon source that operates without electrical energy input, apart from temperature control of the crystal. The researchers point toward satellites, which are bathed in unfiltered sunlight and currently carry lasers for quantum key distribution; a solar-powered source could shed much of the supporting hardware. The same logic applies to scaling quantum computing without adding to the energy burden.
The conceptual point is subtler and arguably more important. The experiment demonstrates that the coherence of a light source in one degree of freedom does not limit the entanglement achievable in another. The team designed the setup so that differences in color and propagation direction did not affect the photons’ polarization, and the theory predicted that if the entanglement lives only in polarization, the pump’s orderliness in its oscillation direction is what matters, not its direction or color. The experiment confirms that prediction with natural light.
The caveats are honest ones. The experiment ran outdoors, so weather and time of day affected the coupled power; the team measured a diurnal variation in pump power, and the 100 to 200 nanowatts delivered to the crystal within the relevant bandwidth are a small fraction of what a laser provides. The detected pair rates are modest, around 1,600 pairs per second per milliwatt of pump, and the efficiency of sunlight utilization is constrained by the crystal’s narrow phase-matching bandwidth. Engineering will be needed to brighten the source, broaden the phase-matching, and turn a proof of principle into a deployable device.
The comparison with laser-pumped sources deserves nuance. Normalized to the pump’s spectral bandwidth, the sunlight-driven pair generation rate is comparable to laser-based SPDC; the difference is that a laser concentrates its power into a narrow line, while sunlight spreads its energy across a broad spectrum, so the absolute photon-pair rate from sunlight is far lower. The team’s achievement is not that sunlight outperforms a laser, but that it works at all, and that the entanglement quality, the property that matters for quantum information, is on par. For applications where power is abundant and free, as it is in space, that trade-off becomes attractive: a satellite’s solar panels already exist, and a concentrator plus crystal could replace a laser and its supporting electronics.
The result also has a historical arc. The same group spent years arguing, against the field’s skepticism, that incoherent light could produce entanglement, first in theory, then with LEDs, and now with sunlight. Each step moved the demonstration from the laboratory toward the world, and the outdoor setup, with its Fresnel lens and solar tracker, is a genuinely different kind of quantum experiment: one that depends on the weather. The result also clarifies the fundamental relationship between coherence and entanglement, showing that the two are decoupled across degrees of freedom in a way that earlier theory did not fully appreciate.
Still, the demonstration closes a chapter of quantum optics that most researchers assumed was closed before it opened. The skepticism the team faced, with prominent figures doubting even photon detection from sunlight-driven down-conversion, gives the result an edge: the physics of entanglement does not require the lab’s most expensive piece of equipment. It can run on daylight.
Sources: Li, C., Brar, J., Küblböck, M., Upham, J., Fattahi, H. & Boyd, R.W. Generating quantum entanglement from sunlight. Optica 13, 1508 (2026). DOI: 10.1364/OPTICA.601797. Preprint: arXiv:2602.15655.