TOPOS · Long-form guide

Quantum Entanglement: entangled photons and Bell's theorem

Quantum entanglement is the engine underneath quantum teleportation. Before a state |ψ⟩ can travel between two labs, those labs need to share a pair of particles whose fates are already linked. This guide explains how entangled photons are made, what Bell's theorem actually proves, and why entanglement — not matter — is what quantum networks move.

What entanglement is (and isn't)

Two particles are entangled when the state of the whole system cannot be written as one particle's state times another's. The canonical example is a pair of photons in the Bell state |Φ⁺⟩ = (|HH⟩ + |VV⟩)/√2: neither photon has a definite polarisation on its own, yet measuring one in any basis instantly fixes what a matching measurement on the other will find.

The result is a correlation stronger than any classical theory allows. It is not a hidden signal — the individual outcomes look completely random on each side. The correlation only becomes visible when the two sides compare their measurement records over a classical channel.

How entangled photons are made

The workhorse source is spontaneous parametric down-conversion (SPDC). A pump laser hits a nonlinear crystal — typically BBO or periodically poled KTP — and, with small probability, a pump photon splits into two lower-energy photons whose polarisations are entangled by phase-matching geometry. SPDC is the source behind almost every long-distance teleportation and Bell-test experiment, from Innsbruck 1997 to the Micius satellite in 2017.

Other platforms produce entangled emissions too: trapped ions (Innsbruck, 2004), nitrogen-vacancy centres in diamond (Delft, 2015), and semiconductor quantum dots. Each trades rate, purity and wavelength differently — the physics of entanglement is the same.

Bell's theorem in one page

In 1964, John Bell asked whether the correlations quantum mechanics predicts could be explained by local hidden variables — pre-existing properties each particle carries, unaffected by measurements on the other. He derived an inequality that any such theory must obey. Quantum mechanics violates it.

Experiments closed the loopholes one by one:

• 1972 — Freedman & Clauser: first violation of a Bell inequality.
• 1982 — Alain Aspect: rapid switching of detector settings.
• 1998 — Weihs et al. (Innsbruck): fully random, spacelike-separated settings.
• 2015 — Hensen (Delft), Giustina (Vienna), Shalm (NIST): loophole-free tests.
• 2022 — Nobel Prize in Physics: Aspect, Clauser, Zeilinger.

The consensus is unambiguous: nature is either non-local or does not have pre-existing values for every measurement. Both options are radical; both are compatible with entanglement being the resource that quantum teleportation consumes.

Why entanglement can't send messages

The no-signalling theorem forbids using entanglement alone to communicate. The marginal statistics on Alice's side are identical whether Bob measures or not, and whichever basis he uses. To turn a correlation into a message, Bob must send classical information — bounded by the speed of light.

This is the point where entanglement meets teleportation: the protocol requires both a shared entangled pair and a classical channel. The pair provides the quantum correlation; the two classical bits provide the instructions that turn the receiver's half into the original state |ψ⟩.

Entanglement as a network resource

A quantum internet is, mechanically, a factory for entangled pairs. Entanglement swapping extends a link: two independent pairs (A–B and B–C) meet at node B, a joint Bell measurement collapses them, and A and C end up entangled without ever having interacted. Chain enough swaps together with quantum memories and you get a quantum repeater — the missing piece between metropolitan links and a continental network.

A short reading map

• Bennett et al. (1993) — the original teleportation protocol using an EPR pair.
• Bouwmeester et al. (1997) — first experimental photonic teleportation.
• Yin et al. (2017) — satellite-based entanglement distribution over 1,200 km.
• Hensen et al. (2015) — loophole-free Bell test in Delft.

FAQ

← Back to TOPOS