Periodic Reporting for period 1 - 2DMultiMems (Two-dimensionally multiplexed on-demand quantum memories)
Reporting period: 2023-10-01 to 2025-09-30
Similar to the classical internet, a future quantum internet has the prospects to significantly impact our daily life. In such an internet, quantum computers will be connected over a network to perform complex computational tasks, e.g. an accurate simulation of drugs or materials. Connection of distant quantum computers will be performed optically through a network of fibers. However, these fibers have intrinsic losses, which limit the maximum distance over which a connection could be established. To connect these quantum processors over intercity distances, quantum repeaters will be dispersed between the quantum computers and entanglement will be distributed over the full network.
A possible implementation of a quantum repeater is the combination of a source of photon pairs and a quantum memory. With this implementation, the source emits entangled pairs of photons of which one is sent to a distant detector and the other is stored in the quantum memory, thus, the quantum memory is entangled with the telecommunication photon. Entanglement is distributed in the network by using two of these quantum repeater nodes with a common detector with Bell-state measurement setup at the distance. A detection event heralds that one of the two memories has absorbed a photon, and since, there is no knowledge on where the photon came from, the quantum memories are in an entangled state.
However, a fundamental limit is imposed on rate at which entanglement will be generated: the quantum repeaters have to wait for the travel time of the photons to the detector and the time necessary for the heralding signal to return before attempting another entanglement trial. Quantum repeaters based on multimode quantum memories overcome this limitation, as these can store entanglement in several degrees of freedom or modes without the limiting waiting time. With multiplexing, the entanglement rate increases linearly with the available number of modes. Quantum memories based on crystals doped with rare-earth-ions promise a particularly high degree of multiplexing, as these quantum memories have the unique prospect of combining time, frequency, and spatial multiplexing in one system.
Until the start of this project, the record of available modes with on-demand storage and retrieval was limited to 30 as only temporal multiplexing was available. The scope of this Marie-Curie project was to explore a new type of quantum memory array that combines spatial and temporal multiplexing for increased number of modes. This project targeted three research objectives: First, we wanted to explore sequences of optical laser pulses as path to increase storage times of the quantum memories. Second, we wanted to store quantum information in the quantum memory array. Thirdly, we wanted to build a second quantum memory array and generate entanglement between these two systems.
A possible implementation of a quantum repeater is the combination of a source of photon pairs and a quantum memory. With this implementation, the source emits entangled pairs of photons of which one is sent to a distant detector and the other is stored in the quantum memory, thus, the quantum memory is entangled with the telecommunication photon. Entanglement is distributed in the network by using two of these quantum repeater nodes with a common detector with Bell-state measurement setup at the distance. A detection event heralds that one of the two memories has absorbed a photon, and since, there is no knowledge on where the photon came from, the quantum memories are in an entangled state.
However, a fundamental limit is imposed on rate at which entanglement will be generated: the quantum repeaters have to wait for the travel time of the photons to the detector and the time necessary for the heralding signal to return before attempting another entanglement trial. Quantum repeaters based on multimode quantum memories overcome this limitation, as these can store entanglement in several degrees of freedom or modes without the limiting waiting time. With multiplexing, the entanglement rate increases linearly with the available number of modes. Quantum memories based on crystals doped with rare-earth-ions promise a particularly high degree of multiplexing, as these quantum memories have the unique prospect of combining time, frequency, and spatial multiplexing in one system.
Until the start of this project, the record of available modes with on-demand storage and retrieval was limited to 30 as only temporal multiplexing was available. The scope of this Marie-Curie project was to explore a new type of quantum memory array that combines spatial and temporal multiplexing for increased number of modes. This project targeted three research objectives: First, we wanted to explore sequences of optical laser pulses as path to increase storage times of the quantum memories. Second, we wanted to store quantum information in the quantum memory array. Thirdly, we wanted to build a second quantum memory array and generate entanglement between these two systems.
First of all, the main outcome of this project is the development of a novel quantum memory array with ten individually-controllable cells. We realize this array with a combination of two acoustic-optic deflectors, two lenses, and a Y2O5 crystal doped with Pr3+. The crystal is located a 3K inside a cryostat between the two lenses and the deflectors, lenses and crystal form a 4f configuration. We implement the quantum memories using the atomic-frequency comb protocol, which natively allows for temporal multiplexing. All cells are prepared simultaneously through a separate optical path with an acoustic-optical deflector. On-demand storage and retrieval are achieved with optical pulses counterpropagating to the preparation. We characterized the efficiencies of the optical system and measure memory efficiencies comparable to systems without spatial multiplexing. With a total of ten memories, we demonstrated quantum storage in 250 spatio-temporal modes using input pulses at the level of single photons. We characterized the cross-talk between the memories and found it to be on average 2.9(1) %. The measured signal-noise-ratio of 10(2) across the array indicate readiness for this system to be connected with a photon-pair source for the distribution of entanglement. Furthermore, we stored qubits encoded in the path or time degree and implemented a tomographic analysis of the retrieved quantum states. The retrieved states showed high fidelities exceeding 90% for both encodings and all ten memories. We then connected the quantum memory array with a photon-pair source and measured non-classical correlations between the quantum memories and telecom photons with an average cross-correlation of 3.3(2) over more than 100 modes. As a step towards entanglement between two quantum memory arrays, we generated entanglement between two on-demand quantum memories with 15 temporal modes.
This Marie-Curie project resulted in several advances beyond the state of the art. This work represents the first solid-state quantum memory array and the demonstrated on-demand storage and retrieval in 250 modes with inputs at the single-photon level significantly exceeds the previous record of 30 modes. We also realized the first storage of path qubits in solid-state quantum memories and simultaneously stored two time-bin qubits in the array. These capabilities open prospects for photonic quantum computing. When connected to a source of photonic cluster states, the quantum memory array may hold these quantum resources until the state has reached a size sufficient for the computational task. This connection requires however further research, as the wavelength of these photonic processors differ from the operating wavelength of the quantum memory array. Additionally, the bandwidth of these photons exceeds significantly the accessible bandwidth of the quantum memories.
In the second year of this project, we entangled two quantum memories with on-demand storage and retrieval. The 15 temporal modes improved the entanglement rate of the system. Up to this demonstration, entanglement of solid-state memories has only been demonstrated without the on-demand capabilities, which are essential for synchronization in long-distance quantum networks.
To advance towards entanglement of two quantum memory arrays, we demonstrated non-classical correlations of the memory array with telecom photons. This on-demand storage harnessed more than 100 spatio-temporal modes and thus significantly surpassed previous demonstrations with only 15 temporal modes. In total, the project resulted in three high-impact publications.
Deployment in a real-world quantum internet requires further research on two aspects. First, the efficiencies of this system have to be improved by integrating an optical cavity with the array. Second, the storage times were limited in these experiments to tens of microseconds. Dynamical decoupling through radio-frequency or optical pulses have to be exploited to increase the storage times. The application of radio-frequency pulses imposes a challenge, as these may act on several memory cells simultaneously, and thus, precise timing of the pulses is required.
In the second year of this project, we entangled two quantum memories with on-demand storage and retrieval. The 15 temporal modes improved the entanglement rate of the system. Up to this demonstration, entanglement of solid-state memories has only been demonstrated without the on-demand capabilities, which are essential for synchronization in long-distance quantum networks.
To advance towards entanglement of two quantum memory arrays, we demonstrated non-classical correlations of the memory array with telecom photons. This on-demand storage harnessed more than 100 spatio-temporal modes and thus significantly surpassed previous demonstrations with only 15 temporal modes. In total, the project resulted in three high-impact publications.
Deployment in a real-world quantum internet requires further research on two aspects. First, the efficiencies of this system have to be improved by integrating an optical cavity with the array. Second, the storage times were limited in these experiments to tens of microseconds. Dynamical decoupling through radio-frequency or optical pulses have to be exploited to increase the storage times. The application of radio-frequency pulses imposes a challenge, as these may act on several memory cells simultaneously, and thus, precise timing of the pulses is required.