The state moves onto a particle already at the receiver
In 1997 Bouwmeester and colleagues brought an input photon together with one photon from an entangled pair. A selected joint detection let them test whether the third photon, already on the receiver’s side, carried the input polarization state. The result realized a successful branch of quantum teleportation: the state was transferred onto a particle already at the destination.
The ideal protocol calls the sender Alice and the receiver Bob. Alice holds an unknown input qubit, a system with two basis states and possible superpositions of them. She also holds half of a pair entangled with Bob. Measuring her two systems together gives a classical label. Bob uses that label to correct his own half, recovering the input state without receiving the original particle.
The protocol uses up the shared entanglement, and the original input no longer survives as an independent perfect copy. The classical label does not reveal the unknown amplitudes. Bennett and colleagues’ 1993 proposal spelled out these resource requirements.
Read the evidence: C. H. Bennett et al. (1993), Teleporting an Unknown Quantum State via Dual Classical and Einstein-Podolsky-Rosen Channels, Phys. Rev. Lett. 70, 1895–1899 — university-hosted copyD. Bouwmeester et al. (1997), Experimental Quantum Teleportation, Nature 390, 575–579 — ETH-hosted copy
Without the label, no readable message
With a maximally entangled pair and a complete Bell measurement, there are four joint outcomes, each paired with a known Pauli correction: a bit flip, a phase flip, both, or neither. Before its label arrives, Bob averages over those outcomes. The left side is Bob’s average density matrix, which predicts his measurement probabilities. Inside the sum, each sigma denotes one of the four Pauli operations applied to the input matrix and paired with its adjoint, the reverse operation. For a normalized input, averaging all four gives half the identity matrix I on Bob’s two-state system.
Without the label, Bob’s state is the same whatever the input was. Sorting the data after the classical message arrives can reveal the transfer, but it cannot make the message arrive sooner. A complete ideal protocol, which sends two classical bits, and an optical experiment that keeps one successful branch work under different conditions.
The 1997 optical experiment selected one branch
Bouwmeester and colleagues brought an input photon and one photon of an entangled pair together at a beam splitter. A suitable pattern of joint detections picked out the antisymmetric Bell-state branch, and the receiver’s photon was analyzed on that condition for several input polarizations.
This optical version succeeded only some of the time: it did not identify and correct all four Bell outcomes on every trial. Testing superposition inputs matters, because a device that only recognizes one preferred polarization basis can mimic a successful transfer in that basis and still fail on other states.
Read the evidence: D. Bouwmeester et al. (1997), Experimental Quantum Teleportation, Nature 390, 575–579 — ETH-hosted copy
Background subtraction and heralding are different analyses
The paper reports two kinds of data. Threefold coincidence curves had a measured spurious contribution subtracted. Fourfold coincidences, which also required detecting the input photon’s partner, were shown for two nonorthogonal inputs without that subtraction. The two analyses select different events.
A visibility is not automatically an unconditional success probability or a process fidelity. Before interpreting a reported value, find its denominator, the losses, what was subtracted and what it was conditioned on. Here verification also detects the output photon. The fourfold events verify selected state transfers; providing an output for a subsequent task requires a different delivery and verification arrangement.
Read the evidence: D. Bouwmeester et al. (1997), Experimental Quantum Teleportation, Nature 390, 575–579 — ETH-hosted copy
Swapping changes who shares entanglement
Pan and colleagues’ 1998 experiment started with two entangled photon pairs. A Bell-state measurement on one photon from each pair conditionally left the two remaining photons entangled. Those two need not have interacted directly.
The claim is about a conditional joint state and how it was verified. It does not mean that either remaining photon carries a locally readable message, or that selecting data later sends a signal into the past. Bell-inequality tests and teleportation benchmarks also ask different statistical questions.
The ideal protocol predicts an exact output state after the matching correction. The optical experiment tested that prediction using selected photon detections, under real source and detector conditions. Source brightness, indistinguishability and detector efficiency determine how often such a successful branch is observed.
The successful detections demonstrate conditional state transfer. Their verification fraction for a fixed input has a narrower meaning than a fidelity characterizing every possible input or an unconditional network success rate. Keeping that distinction makes it possible to appreciate the 1997 result and understand what a working communication network must add.
Read the evidence: J.-W. Pan et al. (1998), Experimental Entanglement Swapping: Entangling Photons That Never Interacted, Phys. Rev. Lett. 80, 3891 — author abstractC. H. Bennett et al. (1993), Teleporting an Unknown Quantum State via Dual Classical and Einstein-Podolsky-Rosen Channels, Phys. Rev. Lett. 70, 1895–1899 — university-hosted copyD. Bouwmeester et al. (1997), Experimental Quantum Teleportation, Nature 390, 575–579 — ETH-hosted copy