Quantum Memory

The goal of the quantum memory program is to realise a memory that will enable quantum networks and computation. For example, quantum key distribution systems are already active over short distances. Long distance key distribution is limited, however, by the losses in optical fibres. An optical quantum memory is an integral part of a repeater device that will extend the range of quantum key distribution. In the longer term, quantum logic operations in optical quantum computers will require optical quantum memory to buffer intermediate results and create deterministic logic gates.

Storage of quantum states of light requires a coherent optical memory. One possible method is to map optical quantum states into atomic ensembles. The Gradient Echo Memory (GEM) is a coherent light storage technique based on photon echoes. The idea was first demonstrated at the ANU in a 2-level solid state ensemble [Hetet et al., Phys. Rev. Lett. 100 023601 2008]

The storage mechanism works as shown in figure 1. An ensemble of two level atoms (a) is frequency shifted (b) to create a gradient of atomic frequencies along the length of the ensemble. The frequency width of the atomic sample is adjusted to capture the entire bandwidth of the optical pulse. After flipping the frequency gradient at time ts, a photon echo is generated at time 2ts (c). The atomic ensemble can be composed of either a two-level atoms (as shown) or three level atoms. In the case of three level atoms, we use an additional control beam to couple atomic ground states, as shown in figure 2. In this way we enable a range of other atomic systems for GEM, in particular the alkali atoms, which have long lived atomic ground states.

The quantum memory program at the ANU has achieved a number of important milestones. In a two-level system, measurements have shown recall beyond the no-cloning limit [Hedges et al., Nature 465 1052 2010]. This limit ensures that it is not possible for an eavesdropper to have reconstructed the quantum state better than the intended recipient.

In the three level system we have demonstrated fully flexible recall of light pulses [Hosseini et al., Nature 461 241 2009]. We can recall pulses in arbitrary order, compress, stretch, split and frequency shift pulses all by manipulating the atomic coherences and the power in the control beam.

In the coming years we will be pursuing higher efficiency memories with longer coherence times by using cold atomic ensembles and more accurate generation of atomic frequency gradients.