Quentin Glorieux, Jeremy Clark, Alberto Marino, Paul Lett

Joint Quantum Institute, National Institute of Standards and Technology and University of Maryland, Gaithersburg, Maryland 20899, USA


The storage of light in a quantum memory (QM) is a crucial tool for the future development of quantum networks. A QM is the key part of a quantum repeater which allows long distance distribution of entanglement. A QM should preserve the quantum state of a light pulse (high fidelity) during the longest storage time possible. In this paper we study a promising technique for the coherent storage and retrieval of light: the gradient echo memory in a warm atomic 85Rb vapor, and we demonstrate that it can be used to store multimode images. A spatially multimode approach for quantum memory can in principle reduce the requirements on the storage time for a realistic implementation of a quantum repeater. As the movement of the atoms during the storage time is a main factor to quantify the spatial resolution of this memory, we present here a study of atomic diffusion for a cell with Ne buffer gas at the pressure of 667 Pa.

Gradient Echo Memory

Gradient echo memory is based on the reversible dephasing of the macroscopic coherence of an atomic ensemble. A pulse of light is sent into a warm atomic 85Rb vapor, which can be considered as a collection of three-level atoms in a Λ configuration. A spatially dependent Zeeman shift is obtained by applying a linearly varying magnetic field (200 mG/cm) across the 5 cm-long rubidium cell. In the presence of a strong control field and for a Zeeman shift larger than the frequency bandwidth of the pulse, each frequency component of the pulse can be absorbed due to the high optical density of the atomic medium. The state of light is then transferred to the long-lived atomic ground state coherence and the spectral components of the signal mapped linearly along the length of the sample. After the excitation, the collective dipole dephases due to the inhomogenous magnetic field. It is possible to recover the ensemble’s macroscopic coherence, however, by reversing the magnetic field gradient and allowing sufficient time for the dipoles to rephase. Upon dipole rephasing, the input light pulse will emerge in the forward direction if the control field is switched on.

The diffusion of the atoms during the storage time will ultimately limit the spatial resolution of our memory. For an image constituted by 3 vertical lines pairs (1951-USAF target) with a given spatial frequency 1/α, the contrast of the retrieved image after a time of storage t is given by:


where C0 is the contrast at t=0 and D is the diffusion coefficient. We present the experimental results for the effect of atomic diffusion on the contrast for four different spatial frequencies as well as the theoretical predictions.


In this paper we have demonstrated that the optical gradient echo memory is suitable for the coherent storage of images. Additionally, we use it to measure the effect of atomic diffusion and show that the maximum spatial frequency to be stored is predetermined by the storage time and the diffusion coefficient. We expect that storing the Fourier transform of the image would increase the maximum spatial frequency achievable. Finally, we would like to emphasize that this setup is perfectly adapted to be combined with recent experiments on the generation of squeezed states and entangled images with four-wave mixing in a hot rubidium vapor.