Physicists in the US are the first to store two entangled quantum states in a memory device and then retrieve the states with their entanglement intact. Their demonstration, which involves "stopping" photons in an ultracold atomic gas, could be an important step towards the practical implementation of quantum computers.

The basic unit of information in a quantum computer is the qubit, which can take the value 0, 1 or—unlike a classical bit — a superposition of 0 and 1 together. A photon could be used as a qubit, for example, with its "up" and "down" polarization states representing 0 or 1.

If many of these qubits are combined or “entangled” together in a quantum computer, they could be processed simultaneously and allow the device to work exponentially faster than its classical counterpart for certain operations. Entanglement could also play an important role in the secure transmission of information because the act of interception would destroy entanglement and reveal the presence of an eavesdropper.

Fragile states

However, this fragile nature of entanglement has so far prevented physicists from making practical quantum information systems.

Now, a team of physicists at the California Institute of Technology led by Jeff Kimble has taken a step towards this goal by working out a way to store two entangled photon states in separate regions of an extremely cold gas of caesium atoms (Nature 452 67).

The entangled states are made by firing a single photon at a beam splitter, in which half the light is deflected left into one beam, and the other half deflected right to form a second beam. These two beams are parallel, separated by about 1 mm and contain a pair of entangled photon states—one state in the left beam and the other in the right.

Slow-moving hologram

Once inside the cloud, a hologram-like imprint of the two entangled photon states on the quantum states of the atoms can be created using an effect called electromagnetically-induced transparency (EIT). This imprint moves through the gas many orders or magnitude more slowly than the speed of light, effectively “stopping” the entangled states for as long as 8 µs.

EIT is initiated by a control laser that is shone through the gas to create the holograms. Then, the laser is switched off, which causes the photon states to vanish leaving the holograms. Then the control laser is switched back on, which recreates the entangled photon states from the holograms. The storage time can be changed by simply varying the time that the laser is off.

When the recreated entangled photon states leave the atomic gas, one of the states passes through a device that shifts its phase, while the other does not. The two states then recombine at a detector. If the states remain entangled, adjusting their relative phase would create a series of bright and dark quantum interference effects at the detector.

20% entangled

By repeating the experiment for a large number of single photons and measuring the interference intensity, the team concluded that about 20% of the entangled photon states were recovered from the atomic gas. While this might seem like a poor success rate, it is good by quantum-computing standards where entanglement efficiencies of 1-2% are common.

According to Lene Hau of Harvard University, who pioneered EIT, the Caltech technique could be improved by cooling the atomic gas below the current 125 mK to create a Bose-Einstein condensate in which all the atoms are in a single coherent quantum state.

Sharing secrets

Hau also believes that the Caltech memory device could be modified to store the entangled states in two different atomic gases. This, she says, would allow quantum keys to be shared securely between users of a quantum encryption system.

The storage and retrieval of individual photons in an atomic gas was first demonstrated in 2005 by two independent groups—one led by Hau and the other working at the Georgia Institute of Technology. The ability to store photons without destroying entanglement is crucial for the transport of single photons over large distances, where the memories would work as quantum repeater that would boost the optical signal without destroying the quantum nature of the signal.