Entanglement swapping is a technique that creates quantum entanglement between two particles that have never met, never interacted, and may be separated by enormous distances. It works by using a third party to perform a specific measurement on two other particles, each already entangled with one of the target particles, effectively transferring the entanglement across to the pair that never touched. The procedure is central to building quantum networks and quantum repeaters, and it raises some of the most striking puzzles in the foundations of physics.
How Two Strangers Become Entangled
Start with two separate pairs of entangled particles. Call them pair AB and pair CD. Particles A and B are entangled with each other, and particles C and D are entangled with each other, but there is no connection whatsoever between the two pairs. Now take the “inner” particles, B and C, bring them together, and perform a joint measurement on them. That measurement destroys the entanglement within each original pair but, in doing so, projects the two “outer” particles, A and D, into an entangled state. A and D are now correlated in exactly the way quantum entanglement predicts, even though they were produced independently and have never been in the same place.
The joint measurement performed on B and C is called a Bell-state measurement. It asks a very specific question about the combined quantum state of those two particles and, depending on the answer, determines which type of entangled state A and D end up sharing. The person holding particle A or D then needs to know the measurement result (sent over a normal classical communication channel) to know exactly how their particles are now correlated. This combination of a quantum measurement and a classical message is the same recipe used in quantum teleportation, and entanglement swapping is sometimes described as teleporting entanglement itself rather than a specific quantum state.
Why the Bell-State Measurement Is the Hard Part
The joint measurement on B and C is the bottleneck of the whole procedure. There are four possible Bell states that B and C can collapse into, and in an ideal world you would be able to distinguish all four. In practice, using only standard optical components like beam splitters and detectors, you can reliably tell apart only two of the four. That caps the success rate of each attempt at 50 percent at best.
This limit is not just an engineering inconvenience. It is a fundamental constraint of linear optics, meaning it persists no matter how good your detectors or how clean your light source. Overcoming it requires either nonlinear optical elements, extra ancillary photons, or matter-based systems that interact with light differently. Recent experimental work has pushed toward breaking this barrier. One approach uses a nanophotonic device that functions as a nonlinear Bell-state analyzer, sidestepping errors caused by multiphoton emission that plague standard linear-optics setups and fundamentally limit both the efficiency and the accuracy of quantum networking protocols.1PubMed. Faithful Quantum Teleportation via a Nanophotonic Nonlinear Bell State Analyzer Another route demonstrated that hyperentanglement, where photons are entangled in more than one property simultaneously, can push the success probability above the 50 percent ceiling even with linear optics.2PubMed Central. Bell-state measurement exceeding 50% success probability with linear optics
The Quantum Repeater Problem
Entanglement swapping exists largely because of a practical headache: quantum signals degrade over distance. A single photon traveling through an optical fiber gets absorbed or scattered, and unlike a classical signal, you cannot simply amplify it. Copying a quantum state is forbidden by the no-cloning theorem. So if you want to share entanglement between two cities hundreds of kilometers apart, you cannot just send a photon and hope for the best.
A quantum repeater solves this by breaking the long link into shorter segments. Each segment generates its own entangled pair, and entanglement swapping at each intermediate node stitches the segments together, extending the entanglement from one end of the chain to the other. Successive swapping operations can, in principle, distribute entanglement over arbitrarily long distances.3Journal of the Optical Society of America B. Quantum repeaters with entangled coherent states Each swap introduces some noise, so the final entangled state is never as clean as the original short-range pairs. But the strategy trades a single impossible long hop for many feasible short hops with manageable degradation.
A key ingredient that makes repeaters practical is quantum memory. Each node needs to hold onto its half of an entangled pair while waiting for the neighboring segment to succeed. Without memory, both segments would need to produce entanglement at exactly the same instant, which is vanishingly unlikely. Experimental demonstrations have paired entanglement swapping with solid-state quantum memories that can absorb and re-emit photons on demand, forming elementary repeater links compatible with multiplexing and high-speed operation.4PubMed. Heralded entanglement distribution between two absorptive quantum memories
Speeding Things Up With Multiplexing
Even with quantum memory, producing one entangled pair per attempt is slow. Every attempt has a chance of failure: the photon might get lost in the fiber, the Bell-state measurement might give an inconclusive result, or the memory might decohere before the classical confirmation message arrives. If you are running one attempt at a time, you spend most of your time waiting.
Multiplexing attacks this by running many attempts in parallel. One recent experiment used a chain of trapped calcium ions as a quantum network node, generating up to 44 time-bin modes per round and sending them through 12 kilometers of fiber. This hybrid multiplexing scheme achieved a heralded ion-photon entanglement rate of about 4.3 events per second over metropolitan-scale distances.5PubMed Central. Metropolitan-scale ion-photon entanglement via a quantum network node with hybrid multiplexing enhancements That rate is modest by classical networking standards but represents a significant step toward making quantum repeater chains practical. The tradeoff is that multiplexing demands more memory slots at each node. Theoretical work on repeater chains has shown that clever memory allocation strategies can keep the average number of memory slots per node constant even as the total distance grows, with entanglement distribution rates that decay only polynomially rather than exponentially with distance.6arXiv. Entanglement Swapping for Repeater Chains with Finite Memory Sizes
Entanglement Swapping Beyond Fiber
Optical fiber works for city-scale links, but intercontinental quantum networks probably need satellites. The idea is straightforward in concept: a satellite acts as the intermediate node, receiving entangled photons from two ground stations (or generating and distributing them), and performing entanglement swapping in orbit. One recent feasibility study analyzed an “uplink” configuration where entangled pairs are generated on the ground and sent up to a satellite for the Bell measurement. This configuration had been dismissed as impractical, but detailed numerical modeling of channel efficiency and pair fidelity suggests it is actually viable for various satellite-station geometries.7Physical Review Research. Quantum entanglement distribution via uplink satellite channels
A complementary approach equips the satellite with quantum memory, turning it into a genuine orbiting repeater node. This lets the satellite store entanglement from one ground station while waiting for a window to connect with the other, relaxing the requirement that both ground links succeed simultaneously.8IEEE International Conference on Quantum Computing and Engineering. Entanglement Swapping in Orbit: a Satellite Quantum Link Case Study Satellite-based repeaters remain in the early design and simulation phase, but they represent the most credible path to global-scale quantum networking.
Different Flavors of Entanglement Swapping
Most introductions to entanglement swapping describe pairs of single photons, but the technique applies to a wider landscape of physical systems. Understanding which platforms work, and where each has advantages, matters for anyone following the field.
Quantum Dots
Semiconductor quantum dots can produce entangled photon pairs on demand rather than probabilistically. Two groups independently demonstrated entanglement swapping using photons from gallium arsenide quantum dots in 2019. One experiment achieved a Bell-state fidelity of about 0.81 after swapping, enough to violate Bell inequalities and confirm genuine entanglement in the output.9PubMed. Entanglement Swapping with Semiconductor-Generated Photons Violates Bell’s Inequality The other showed that these on-demand sources could perform the full swapping protocol without needing the spectral or temporal filtering that spontaneous sources typically require.10PubMed. Entanglement Swapping with Photons Generated on Demand by a Quantum Dot Deterministic sources like quantum dots could eventually simplify repeater architectures because you know when each photon pair is coming, rather than having to herald success after the fact.
Continuous-Variable Systems
Instead of encoding quantum information in the polarization or presence of individual photons, continuous-variable approaches use properties like the amplitude and phase of a light field, which can take on a continuous range of values. Entanglement swapping works here too, though the physics looks quite different. Instead of counting single photons, you perform homodyne measurements on squeezed light beams. The challenge is that continuous-variable entanglement is more fragile in noisy environments. Theoretical analyses of entanglement swapping with squeezed states coupled to thermal noise have mapped out the conditions under which the output state remains genuinely inseparable.11Chinese Physics. Continuous variable entanglement swapping in a thermal environment An all-optical entanglement swapping scheme has also been demonstrated that works in both discrete and continuous-variable systems, underscoring that the swapping concept is not tied to any single encoding.12PubMed. All-Optical Entanglement Swapping
Hybrid Networks
Real-world quantum networks will likely combine different platforms: maybe a trapped-ion processor at one end, a photonic link in the middle, and a solid-state memory at the other. The problem is that these platforms naturally produce different types of entangled states. Hybrid entanglement swapping addresses this by performing a Bell-state measurement that bridges particle-like and wave-like optical qubits, creating entanglement between nodes that would otherwise be incompatible. An experimental demonstration of this approach generated single-photon entanglement and hybrid entanglement and then used a tailored Bell measurement to produce hybrid entanglement at a distance.13PubMed Central. Connecting heterogeneous quantum networks by hybrid entanglement swapping This kind of interoperability is essential if quantum networks are going to grow beyond isolated lab demonstrations into modular, scalable infrastructure.
Noise, Loss, and Purification
Every real entanglement swapping operation is imperfect. Photon loss in the channel, dark counts in detectors, imperfect Bell-state measurements, and decoherence in quantum memories all conspire to reduce the quality of the resulting entangled state. After several rounds of swapping in a repeater chain, the accumulated noise can destroy any useful entanglement.
The standard remedy is entanglement purification (also called distillation). You sacrifice some number of noisy entangled pairs to produce a smaller number of higher-quality pairs. Purification protocols typically involve both parties performing local operations on their halves of multiple pairs and then comparing certain measurement results over a classical channel. Pairs that pass the check are kept; those that fail are discarded. Demonstrations on real quantum hardware have shown that purification can restore entanglement quality after swapping through a noisy channel that includes combined bit-flip, phase-flip, and phase-change errors.14Springer. Demonstration of entanglement purification and swapping protocol to design quantum repeater in IBM quantum computer In a full repeater chain, purification and swapping alternate: you purify the elementary links, swap to extend them, purify the extended links, swap again, and so on.
Reducing Delay in Multi-Hop Chains
In a long repeater chain, the naive approach is sequential: swap segment 1 with segment 2, then swap the result with segment 3, and so on down the line. Each swap requires waiting for a classical message confirming success before the next one can proceed. This sequential delay grows linearly with the number of nodes, and for intercontinental distances with many hops, it becomes a serious performance drag.
Simultaneous entanglement swapping flips the strategy. Instead of chaining swaps one after another, all intermediate nodes perform their Bell measurements at roughly the same time. The classical results are broadcast in parallel, and the end nodes use the combined information to determine their shared state. This cuts the delay dramatically. A hierarchical version goes further by organizing the swaps into levels: a first round of simultaneous swaps within local groups, followed by a second round that stitches the groups together, reducing not just the latency but also the total amount of classical information that needs to be transmitted to any single node.15Chinese Physics B. Hierarchical simultaneous entanglement swapping for multi-hop quantum communication based on multi-particle entangled states
Loophole-Free Bell Tests and the Foundations of Physics
Entanglement swapping has also become a tool for probing the deepest questions about quantum mechanics. One of the most important is whether the correlations predicted by quantum theory could be explained by some hidden local mechanism. Bell tests are designed to rule out such explanations, but every practical Bell test has potential loopholes. Maybe the detectors are too inefficient. Maybe the measurements are not truly random. Maybe the particles had time to exchange a signal.
Entanglement swapping offers an elegant way to close multiple loopholes simultaneously. In a landmark experiment, researchers used swapping to entangle two rubidium atoms separated by 398 meters, then measured their spin states fast enough and efficiently enough to close both the detection loophole and the locality loophole in a single experiment.16PubMed. Event-Ready Bell Test Using Entangled Atoms Simultaneously Closing Detection and Locality Loopholes The “event-ready” feature is key: the heralding signal from the entanglement swapping step tells you when a valid pair has been created before you measure it, so you avoid post-selecting only the events that happened to work. This type of experiment was proposed as a route to loophole-free Bell tests precisely because entanglement swapping between atom-photon systems separates the entanglement creation from the measurements in a way that is hard to achieve with photons alone.17New Journal of Physics. Heralded mapping of photonic entanglement into single atoms in free space: proposal for a loophole-free Bell test
The Delayed-Choice Puzzle
Perhaps the most mind-bending aspect of entanglement swapping is what happens when you delay the Bell-state measurement. In the standard scenario, particles B and C are measured before anyone looks at particles A and D. But what if you measure A and D first, record the results, and only afterward decide whether to perform the Bell-state measurement on B and C? Experiments have shown that the correlations between A and D depend on a choice that, from a temporal standpoint, has not been made yet when A and D are detected.
This “delayed-choice entanglement swapping” scenario has prompted significant philosophical debate. Some interpretations suggest that the future measurement retroactively determines the past entanglement status of A and D. Others argue that no retrocausation is needed if you interpret the quantum state carefully. One philosophical analysis proposes that the results are naturally explained by countenancing “timelike entanglement,” where the entanglement relation extends across time as well as space, offering a fully realist explanation without requiring particles to somehow know what will happen to them later.18Studies in History and Philosophy of Science Part B: Studies in History and Philosophy of Modern Physics. Timelike entanglement for delayed-choice entanglement swapping Whatever interpretation one favors, the experimental facts are not in dispute: the correlations appear regardless of the temporal ordering of the measurements. Entanglement swapping in its delayed-choice version remains one of the starkest demonstrations that quantum mechanics defies our everyday intuitions about cause and effect.