Porteur : Kastberg Anders
Quantum information relies on the fundamental laws of quantum physics to provide a new way, theoretically more secure and more efficient, to communicate and process information. Here, quantum bits (or qubits), made of coherent superpositions of states and/or entanglement, are the basic information units or resources. On one hand, quantum communication aims at distributing qubits, and quantum cryptography, its major application, offers a provably secure way to establish secret keys between distant partners. On the other hand, quantum computation takes advantage of new algorithms to solve complex problems, such as the factorization of large numbers, in highly reduced computing times compared to the classical case. Qubits can in principle be carried by any quantum system (photons, atoms, ions, etc.) provided it exhibits a two-level observable (polarization, emission times, energy levels, spins...). In this context, photons are considered as the natural “flying qubit” carriers for distribution, and atoms or ions are rather seen to be suitable for storing and computing tasks. In principle, light-matter interaction should allow switching from one type of carrier to another, but a very few proof-of-principle experiments have been reported to date, essentially because of significant mismatches in the carriers operation wavelengths and linewidths. More specifically, in long-distance quantum communication, the carrier photons are usually at the telecom wavelength of 1550 nm, associated to a bandwidth of about 100 GHz (≈ 1 nm). On the other hand, expected storage systems based on cold atomic ensembles or rare-earth ions embedded in crystalline matrices rather exhibit transitions below 900 nm with a typical linewidth ranging from a few MHz to 1 GHz (≈ 100 fm to 10 pm).
Despite this, future quantum information applications will likely rely on the coupling between these different types of qubits, one of the main goals being the realization of quantum networks made of interconnected nodes where storage and/or computation tasks are possible. Such a node, known as a quantum repeater, necessitates the integration and the coupling of several optical components and/or functions, in particular sources of entangled photons operating at telecommunication wavelengths, propagation through optical fibers, quantum interfaces for appropriate coherent wavelength adaptation, quantum memories, full entangled (or Bell) state analysis, and entanglement purification. The implementation of such an embryonic quantum network constitutes a milestone of primary importance for new generation quantum communications.
More specifically, the work we aim at developing concerns the study and the realization of quantum storage of two polarization entangled photons in two distinct quantum memories, one made of cold rubidium atoms and the other made of cold cesium atoms. The photon pairs will be emitted at the telecom wavelength of 1560 nm using a home-made non-linear integrated optics generator. By employing an additional fiber-pigtailed micro-cavity with appropriate finesse, or recently released fiber Bragg grating filters, the associated bandwidth is expected to be below 100 MHz so as to match the dedicated quantum memory transitions. Then, two quantum interfaces, one for each photon, made of identical non-linear integrated optics frequency converters will be used to coherently transpose their initial telecom wavelength to those of the atomic ensembles, i.e., 780 nm for the rubidium and 852 nm for the cesium ensembles, respectively.
For both atomic ensembles, we have the possibility to exploit all optically trapped cold atomic samples, in a dilute thermal state and/or in the form of Bose-Einstein condensates (BEC). The right storage protocol will have to be defined, but most probably the quantum state onto long-lived spin waves, by a spontaneous Raman scattering process. In order to transfer a qubit encoded to the photon in the shape of polarization to the atomic sample, we will probably use either an interference scheme involving the geometric phase, or the Zeeman degeneracy that is present in both Rb and in Cs.
Dans la même rubrique :