The challenge of long-distance quantum communication stems from the inherent loss of photons in optical fibers. Quantum repeaters, designed to overcome this, face significant hurdles due to their system complexity. Transportable quantum memories and quantum-memory-equipped satellites offer alternative approaches, but require optical quantum memories with lifetimes on the order of hours for global-scale communication. Currently, the longest demonstrated optical storage time is approximately one minute. This research focuses on achieving significantly longer optical storage times using a novel approach leveraging the long spin coherence times demonstrated in solid-state systems. Specifically, we aim to address the challenges of maintaining optical coherence in a zero-first-order-Zeeman (ZEFOZ) magnetic field and to overcome the limitations of reduced effective absorption in these fields. The long spin coherence lifetimes, achieved previously through the application of dynamical decoupling, present an opportunity to extend optical storage times. This work seeks to bridge the gap between long-lived spin coherence and long-lived optical storage by using an atomic frequency comb (AFC) memory protocol within a ZEFOZ magnetic field.

Previous research has demonstrated impressive spin coherence lifetimes—up to six hours in europium-doped yttrium orthosilicate (Eu3+:Y2SiO5). However, translating this spin coherence into long-lived optical storage remains a challenge. The complex energy level structures in ZEFOZ fields and reduced absorption in these fields have hindered progress. Existing optical storage methods achieve maximum storage times of around one minute in 87Rb atoms and a Pr3+:Y2SiO5 crystal using electromagnetically induced transparency, while single-photon level storage currently reaches only about one second. The atomic frequency comb (AFC) protocol, successful in spin-wave storage of photonic qubits, has demonstrated improvements, such as 0.5 s storage in Eu3+:Y2SiO5; however, this still falls short of the requirements for global quantum communication.

This study utilizes a spin-wave based optical storage protocol, employing the atomic frequency comb (AFC) method within a ZEFOZ magnetic field (ZEFOZ-AFC). To maximize effective absorption, an isotopically enriched 151Eu3+:Y2SiO5 crystal with a doping level of 1000 ppm is used. The experimental setup involves a superconducting magnet generating a 1.280 T ZEFOZ field, along with Helmholtz coils for implementing dynamical decoupling (DD) sequences. A laser locked to an ultra-stable cavity provides the pump and probe beams. A detailed energy level characterization is first performed using continuous-wave and pulsed Raman heterodyne detection (RHD) to accurately determine the energy levels in the ZEFOZ field. The optical storage process begins with spectral hole burning, class-cleaning to select a single class of ions, and spin polarization. Then, an AFC is created in the selected energy levels. A probe pulse is absorbed, transferred into a spin-wave excitation by a control pulse, and stored. Dynamical decoupling (DD) sequences, specifically Carr-Purcell-Meiboom-Gill (CPMG) and concatenated dynamical decoupling (KDD), are employed to protect the spin coherence during storage. Finally, a second control pulse transfers the spin-wave back to an optical excitation, resulting in an AFC echo. The coherent nature of the storage is verified using a time-bin-like interference experiment, measuring the visibility of interference fringes after varying storage times. Storage efficiency is analyzed considering the AFC efficiency, control pulse transfer efficiency, and spin storage efficiency. The effects of DD sequence parameters and RF coil heating on the storage efficiency are also investigated.

The experiment achieved a coherent optical storage time exceeding one hour (specifically, the longest storage time of 52.9 ± 1.2 minutes is obtained using the CPMG sequence with a pulse interval of 100 ms). This represents a ~6000-fold improvement compared to the previous state-of-the-art results. The coherent nature of the storage was verified via a time-bin-like interference experiment showing high visibility (93.0 ± 5.5% at 5 minutes and 92.9 ± 4.9% at 60 minutes, corresponding to fidelities of 96.5 ± 2.8% and 96.4 ± 2.5%, respectively) even after one hour of storage. The observed storage time of more than 1 hour is in agreement with the spin coherence time measured independently using RHD. The total storage efficiency was analyzed, revealing components like AFC efficiency (ηAFC ≈ 2.5% after DD), control pulse transfer efficiency (ηcontrol ≈ 38.5%), and spin storage efficiency (ηspin up to 14.1% with KDD for 5 minutes storage). The limitations in storage efficiency were mainly attributed to insufficient pulse bandwidth relative to spin inhomogeneous broadening and inhomogeneities in the RF field produced by the coils. The time-bandwidth product of the memory is estimated as 3.6 × 10⁹.

The results demonstrate the feasibility of achieving hour-long coherent optical storage using a combination of ZEFOZ techniques, AFC protocols, and dynamical decoupling. This significant increase in storage time pushes the boundaries of what's possible in quantum memory technology, significantly advancing the potential for large-scale quantum communication networks. The achieved fidelity in the time-bin-like interference experiments further solidifies the suitability of this approach for encoding and preserving quantum information. While the current efficiency is still relatively low, there are clear paths to improvement, such as enhancing the AFC efficiency through cavity enhancement techniques, improving the control pulse transfer efficiency by optimizing the crystal orientation, and enhancing the homogeneity of the RF field. The presented work serves as a substantial step towards realizing a practical, long-lived quantum memory for global quantum communication.

This research successfully demonstrated a coherent optical storage time exceeding one hour using a 151Eu3+:Y2SiO5 crystal with a ZEFOZ-AFC memory protocol and dynamical decoupling, achieving a significant improvement compared to previous results. This breakthrough opens exciting possibilities for developing scalable quantum communication networks. Future research should focus on improving the storage efficiency for single-photon applications and exploring the scalability of this method for larger quantum networks.

The current implementation suffers from relatively low total storage efficiency, primarily due to limitations in AFC efficiency, control pulse transfer efficiency, and the inhomogeneous broadening of the spin transitions. Improving the homogeneity of the RF field and addressing the heating effects of the DD pulses will be crucial in future iterations. Further improvements to the optical pumping scheme are needed to reduce noise sources. Also, scaling this method to a multi-mode memory requires further investigation.