Quantum communication founds on the possibility to encode quantum states into photons and allows, for instance, the provable secure distribution of encryption keys. Despite a lot of progress during the past two decades, quantum communication still faces a distance barrier, which can be overcome by quantum repeaters whose currently most challenging ingredient is a memory for quantum states of light. In addition to robustness and ease of use, the requirements for such a quantum memory include high fidelity quantum state storage and the preservation of entanglement, sufficiently long storage time with on-demand recall, high efficiency, large bandwidth, and high multimode capacity. Six years ago, some of these properties had already been demonstrated using atomic vapor. However, practical solid-state memories were virtually unexplored in view of quantum state storage. The main goal of this thesis was to realize a solid-state quantum memory for light with emphasis on meeting requirements for long distance quantum communication. First, we identified a potential material candidate, a thulium doped lithium niobate waveguide, and conducted spectroscopic investigations to assess its suitability for quantum state storage using a photon-echo type approach. Next, we demonstrated that our storage device is suitable for high fidelity quantum state storage and for the storage of photonic entanglement, and that its large spectral multimode capacity allows for on-demand selective recall in conjunction with standard spectral filters and frequency shifters. In addition, by performing two-photon interference experiments and Bell-State measurements, we found that our memory preserves not only quantum information, but, more generally, the entire photonic wave function, which further confirms its suitability for quantum repeaters as well as for linear optics quantum computers. Finally, we showed that our integrated device also allows for general temporal and spectral manipulation of individual quantum optical pulses, which paves the way towards on-chip quantum optical processors. While more work remains to be done, in particular to improve memory efficiency and storage time, the large number of achievements, together with known ways to overcome the remaining obstacles, makes us confident that a quantum memory suitable for quantum repeaters will soon be built.
Quantum communication founds on the possibility to encode quantum states into photons and allows, for instance, the provable secure distribution of encryption keys. Despite a lot of progress during the past two decades, quantum communication still faces a distance barrier, which can be overcome by quantum repeaters whose currently most challenging ingredient is a memory for quantum states of light. In addition to robustness and ease of use, the requirements for such a quantum memory include high fidelity quantum state storage and the preservation of entanglement, sufficiently long storage time with on-demand recall, high efficiency, large bandwidth, and high multimode capacity. Six years ago, some of these properties had already been demonstrated using atomic vapor. However, practical solid-state memories were virtually unexplored in view of quantum state storage. The main goal of this thesis was to realize a solid-state quantum memory for light with emphasis on meeting requirements for long distance quantum communication. First, we identified a potential material candidate, a thulium doped lithium niobate waveguide, and conducted spectroscopic investigations to assess its suitability for quantum state storage using a photon-echo type approach. Next, we demonstrated that our storage device is suitable for high fidelity quantum state storage and for the storage of photonic entanglement, and that its large spectral multimode capacity allows for on-demand selective recall in conjunction with standard spectral filters and frequency shifters. In addition, by performing two-photon interference experiments and Bell-State measurements, we found that our memory preserves not only quantum information, but, more generally, the entire photonic wave function, which further confirms its suitability for quantum repeaters as well as for linear optics quantum computers. Finally, we showed that our integrated device also allows for general temporal and spectral manipulation of individual quantum optical pulses, which paves the way towards on-chip quantum optical processors. While more work remains to be done, in particular to improve memory efficiency and storage time, the large number of achievements, together with known ways to overcome the remaining obstacles, makes us confident that a quantum memory suitable for quantum repeaters will soon be built.
The ability of storing and retrieving quantum states of light is an important experimental challenge in quantum information science. A powerful quantum memory for light is required in a quantum repeater, which would allow long distance (>500km) quantum communications. To be implemented in such application, the quantum memory must allow on-demand readout, with high fidelity and efficiency, and a long storage time. Additionally a multimode capacity (for temporal or spatial modes) would allow multiplexing. Our approaches focus on rare-earth doped crystals, i.e. solid state quantum memory. I present, in this work, our contributions for a solid-state quantum storage, with good performances in every criteria. In particular, I present the preservation of quantum entanglement during the storage, which paves the way for the implementation of quantum memories in quantum repeaters.
"Quantum communication is the art of transferring a quantum state from one place to another". The capability of distributing quantum states allows, for instance, provably secure communication and, in the near future, will enable the interconnection of quantum computers to establish a quantum internet. A convenient way to send quantum states from one place to another is to encode them into photons and distribute them using optical bres. The distance over which this can be achieved is still limited to a few hundred of kilometers. However, it can be extended in principle to arbitrary long distances by means of quantum repeaters. The realization of such quantum repeaters will require the engineering and interfacing of heterogeneous elements such as entangled photon-pair sources, quantum memories and entanglement swapping operations. The main goal of this thesis is to develop and study the compatibility of the main elements required to build a frequency multiplexed quantum repeater. This architecture requires a frequency multimode entangled photon pair source which, for example, could be implemented exploiting the process that involves the interaction of a strong light pulse with a non-linear crystal. Such a process is versatile and allows easy tailoring of the photon-pair spectral properties in order to match the requirements of the frequency multiplexed quantum repeater architecture. First, as a precursor of such a photon pair source, we developed a novel heralded single photon source that involves spectrally resolved single photon detection and on-demand frequency shifting. These capabilities are also essential tools in a frequency multiplexed quantum repeater. Next, we demonstrated the compatibility of such a sources with frequency multimode photonic quantum memories by performing spectrally multiplexed storage of single photons. We also tested the capability of high- delity storage of photons encoding quantum information in their polarization degree of freedom. Finally, we showed that entanglement swapping operations, i.e. Bell-state measurements, can be performed using photon pair sources compatible with quantum memories, quantum repeater architectures, and real-world scenarios. While a lot of e ort remains to be done, in the work presented in this thesis we demonstrated that all the key elements required to build a frequency multiplexed quantum repeater can be built and eventually be interfaced. The main obstacles towards that goal and possible ways to overcome them are also discussed.
Advanced applications of quantum information science, such as long-distance quantum communication based on quantum repeaters, require photons to be interfaced with different devices. These devices, such as single photon detectors, quantum memories, quantum signal processors, and linear optical elements, may be integrated on a single chip to enable efficient, robust and scalable implementations. With aim of constructing a quantum repeater, and towards realizing integrated optical quantum memories and signal processors in general, we experimentally develop quantum light-matter interfaces using a cryogenically-cooled rare-earth-ion-doped waveguide, namely thulium-doped lithium niobate (Tm3+:LiNbO3). As the basis for our work, we describe a quantum repeater architecture that uses spectral multiplexing and quantum memories that operate with a fixed storage time. Our repeater architecture promises efficient operation, is compatible with the properties of cryogenically-cooled rare-earth-ion-doped crystals, and simplifies the demands of quantum memories compared to other architectures. To demonstrate the feasibility of our repeater design, we utilize our light-matter interface as a quantum memory to perform several experimental demonstrations. These include high-fidelity storage and retrieval of 26 spectrally-multiplexed quantum bits encoded into single-photon-level laser pulses, high-fidelity storage and retrieval of single and entangled photons derived from a photon-pair source, and high-visibility two-photon interference between weak laser pulses that are stored in one or two, separate, quantum memories. The unique properties of our light-matter interface allow us to demonstrate additional protocols that further highlight its promise for integrated quantum signal processing. By combining our Tm3+:LiNbO3 waveguide with a LiNbO3 waveguide phase modulator, we demonstrate operations such as sequencing, filtering, interference, and compression of laser pulses attenuated to the single-photon level. In another experiment, we measure the cross-phase modulation between strong pulses, by which we demonstrate photon-photon interactions that point towards non-destructive measurements of quantum bits. Since the performance of all aforementioned protocols, and future light-matter applications, hinge on the basic spectroscopic properties of our waveguide, we measure these properties at temperatures as low as 800 millikelvin. Our findings, discussed over several attached publications, usher the development of integrated quantum memories and processors for multi-component optical quantum information applications such as repeaters and networks.
This thesis demonstrates an efficient quantum memory that for the first time preserves more quantum information than it loses. Beyond this, it describes how even higher efficiency may be realised without sacrificing the other performance characteristics that are required for a practical device. A quantum memory is a device to store and recall the quantum state of light. Such a memory will be required for most applications of the emerging field of quantum information. The closest practical application is quantum repeaters for long distance encryption. A more distant target is quantum computing. Yet demonstrations of quantum memory to date fall far short of the performance required for repeaters, and further short of that for quantum computing. For these applications, high performance will be required in terms of 3 parameters: long storage time, high efficiency, and a large mode-capacity. Prior to this work, the storage time was the only paramter to have been demonstrated near the required level[1], and that only in a classical regime. The primary result of this thesis is to demonst rate a memory with an efficiency close the required level (70% is measured), and confirm its quantum nature with noise measurements. This represents the first quantum memory with efficiency greater than 50% and the first to operate unconditionally above the "no-cloning limit". The demonstration used the same storage medium as for the previous long-term storage experiment (Pr{u00B3}{u207A}:Y{u2082}Si0{u2085} (PrYSO)), and could in principle reach the same long storage times. The technique used is an optical "gradient echo", an optical equivalent to the long studied gradient echo technique used in Nuclear Magnetic Resonance (NMR). Aside from it's practicality for demonstrations, a gradient-echo also offers a powerful handle for manipulating stored light. However it is more complex than related memory techniques because it employs a spatially-varying medium. This becomes crucial when the optical depth (and memory efficiency) is increased; propagation effects mean the light is stored very differently than in other techniques. A simple linear-response model is used to gain insight into this high optical depth regime. As a result, some new properties are identified which may increase the efficiency in practical circumstances and improve flexibility. Of particular note, it's found that the space-dependence can be used to hide one narrow transition with another. This allows the somewhat unintuitive notion that a light pulse can fully traverse a medium without exciting a transition that it is resonant with. The spectral-hiding property has promise to overcome the problem with simultaneously acheiving all three parameters required for quantum memories. The problem is related to the preparation of available rare-earth ion doped materials. To acheive high efficiency and storage times, these require spectral holeburning whose bandwidth is limited by the splittings of spin levels. Methods exist to create high bandwidth delays lines with holeburning[2], but they cannot be efficient or long-lived. By utilising the spectral-hiding property of the gradient echo, methods to overcome this limitation are described for the first time.
(English) Future networks for quantum communication will require a new form of technology in order to operate over long distances. The no-cloning theorem precludes the use of amplification, so quantum repeater schemes have been proposed to achieve long-distance quantum communication through the distribution of entanglement. A key device for this setup is a quantum memory, which is required in order to store entanglement and synchronise its distribution. Among the many physical systems than can be used as a quantum memory, rare-earth ion-doped crystals stand out due to their long coherence times and potential for multiplexing and integration. This thesis reports the use of one such system, Pr3+:Y2SiO5, as a quantum memory using the atomic frequency comb (AFC) protocol. Building on previous work, we improved the performance of the memory to be able to demonstrate light-matter entanglement with a telecommunications wavelength photon and the quantum memory, for both a bulk memory and an integrated version. To measure light-matter entanglement, we use photon pairs generated through cavity-enhanced spontaneous parametric down conversion, where one photon (the signal) is stored in the memory and the other (the idler) is at a telecommunications wavelength. These photon pairs exhibit energy-time entanglement, which we analyse in the time basis using fibre- and AFC-based Mach-Zehnder interferometers. In the first experiment, we demonstrate light-matter entanglement where the bulk quantum memory is used with on-demand retrieval via storage in a spin-wave. We measure interference fringes with average visibilities of 89(2)% for AFC storage, and 70(3)% for spin-wave storage, sufficient to demonstrate entanglement. This is the first such demonstration using an on-demand multimode solid-state memory and telecom photon. We show that entanglement is maintained for up to 47.7 µs of storage. These results were made possible through improvements in the AFC efficiency and the noise filtering. Furthermore, we perform on-demand storage and retrieval of up to 30 temporal modes. We take the next step towards implementing our system in a real world scenario by measuring entanglement over a distance. To do so, we send the idler photons through both optical fibre spools and deployed fibre in the metropolitan area of Barcelona, Spain, while the photon detection still occurs in the same lab. The resulting visibilities (80% or higher) demonstrate that the photonic qubit does not decohere during transmission through the optical fibre. We then decouple the photon creation and detection by measuring non-classical correlations between signal photons detected in the lab, and idler photons detected 16 km away (44 km in fibre) in a different location. The final experiment uses a fibre-integrated memory for light-matter entanglement. We use a Pr3+:Y2SiO5 crystal containing a Type I laser-written waveguide, interfaced directly with optical fibre. The fibre-integration allows for improved transmission and AFC efficiency compared to a free-space coupled waveguide memory. To analyse the performance of the memory, we demonstrate non-classical correlations for storage times up to 28 µs. We perform qubit tomography in the time basis for storage times of 3 and 10 µs, measuring two-qubit fidelities of 86(2)% and 86(4)%, respectively, demonstrating storage of entanglement. These experiments demonstrate the potential of our system for use as part of a quantum repeater, with many opportunities for further improvement to the storage time, efficiency, and multiplexing capability.
As the reach of quantum technologies extends ever further in communication and information science, a reliable way of transferring quantum information between distant locations becomes ever more crucial. While photons are widely accepted as excellent carriers due to their speed and low decoherence, losses of transmission (in free space or fibre) and the impossibility of cloning quantum information still pose a great challenge. The quantum repeater architecture was suggested as a solution to both problems [1]. In a quantum repeater the information encoded in an input state is transferred to a new one through entanglement swapping, that is then sent on along the channel. In this work we present our advances towards the realisation of a quantum repeater. Our system of choice combines a solid-state quantum memory with a source of photon pairs. The memory is based on a Rare-Earth Doped crystal, where quantum information can be stored in Pr3+ ions as a collective excitation using the Atomic Frequency Comb technique. On demand retrieval of the information is realised by transferring the excitation to a long-lived spin state. Record values of storage times and retrieval efficiencies have been demonstrated in this system [2]. Entangled pairs of single photons are generated by parametric down conversion in a periodically poled crystal placed inside a bow-tie cavity. This allows us to generate narrow band photons pairs, where the signal is spectrally matched to the memory (606nm), while the idler is in the telecom band [3]. Such a configuration allows us to benefit from the high performance of the memory, that also allows for temporal [2] and frequency [4] multimodality, while at the same time overcoming the high optical losses of 606nm photons by pair generation of a telecom photon. The first stepping stone, progress towards which is presented in this work, is the successful demonstration of energy-time entanglement between the telecom idler photon and the signal photon, stored as spin-wave excitation. The entanglement of the original pair is maintained by the memory temporal multimodality. The entanglement analysis will be made through time-bin qubit analysers made of a fibre-based Mach-Zehnder interferometer, for the former, and a solid-state equivalent based on two AFC with different storage times, for the latter [5]. In this direction we have already doubled the efficiency of the AFC storage protocols, that will be beneficial to count rates and signal-to-noise ratio. With respect to [2], we also increased the spectral-matching between the source and the memory [4]. Our experiment will provide an increase in storage time of 3 orders of magnitude with respect to previous demonstrations, as well as introducing for the first time on-demand read-out in a highly multi-mode memory. Demonstration of the successful transfer of quantum information between the signal photon and the long-lived solid-state excitation will open the way to the demonstration of long-distance entanglement between individual nodes in a quantum network.