Performance of a Hot Rb Vapour-Based Portable Quantum Memory

AI summary

Overview

• Research area: Experimental quantum communications — atomic-ensemble quantum memories (AMO physics) characterised for use in quantum repeater networks.
• Technical level: Intermediate. The paper is an experimental characterisation study; readers need some familiarity with optics, atomic physics, and telecom metrics (SNR, BER), but the core logic is accessible.
• Scope (one sentence): The paper measures the coherence time, absolute storage efficiency, signal-to-noise ratio and bit error ratio of a commercial, rack-mountable warm-rubidium-vapour quantum memory across vapour temperatures from 60 °C to 90 °C, and compares the results against the benchmarks used in classical optical communications.

What This Paper Is About

Quantum repeaters are needed because light lost in optical fibre cannot be amplified for quantum states (the no-cloning theorem), and a practical repeater node needs at least two quantum memories. The authors ask whether a commercially available, portable memory — the Qunnect Qu-Mem, built on a warm vapour of ⁸⁷Rb atoms — performs well enough to slot into existing fibre communication architectures. Their goal is to express the device's performance in the same language telecom engineers already use (SNR in dB, BER thresholds) rather than only in quantum-optics terms.

Key Contributions

1. First characterisation of a portable commercial quantum memory. The authors describe the device as "a first-of-its-kind, portable, and scalable QMem device" and measure it as a complete, rack-mountable system rather than a laboratory breadboard.
2. Temperature-resolved performance mapping. Coherence time, absolute efficiency, SNR and BER are measured over a readily controllable vapour temperature range of 60 °C to 90 °C, revealing that different metrics peak at different temperatures.
3. Translation of quantum-memory metrics into classical telecom conventions. Efficiency is defined as a ratio of integrated retrieved to input pulse energy, SNR is computed from time-averaged signal over detector RMS noise, and BER is derived from the measured SNR via the Q-function, allowing direct comparison with classical optical link standards.
4. A polarisation-agnostic optical design. The system uses polarisation beam splitters in a Sagnac configuration, polarisation rotation of the vertical rail, conversion to circular polarisation before the cell, and a Glan-Taylor calcite polarisation filter, so it can accept qubits encoded in any arbitrary polarisation.

Main Findings

• Coherence time falls as temperature rises. The 1/e coherence time of optical storage decreased with increasing vapour temperature. At 60 °C the measured coherence time was 278 ± 18 μs. Long-storage-time traces contained an extra classical offset, attributed to ground-state decoherence producing photons incoherent with the stored probe; subtracting this offset gave coherence times "similar to other studies."
• Efficiency is non-monotonic in temperature. Storage efficiency peaked near 70 °C. A second investigation over a larger temperature range, calibrated to the vapour-cell sensor reading rather than the set-point temperature, confirmed the same trend. The rise is attributed to higher optical depth and greater atom-photon interaction probability; the later decline is attributed to re-absorption of the probe before it exits the cell. The probe transmission through the vapour-cell system was 0.75.
• SNR is highest at the top of the tested range. The maximum SNR occurred at 90 °C, the highest temperature tested. At a 5 μs storage time the system reached 10.9 dB to 16.8 dB; for storage times near the 1/e coherence time at each temperature, SNR ranged from 8.7 dB to 12.3 dB. Classical protocols are cited as requiring 10 dB to 25 dB, so the short-storage-time values fall inside that band. The paper states the system yields SNR values sufficient to serve as a classical optical channel for storage times between 100–200 μs.
• BER meets telecom thresholds at most temperatures. Using BER = Q(√SNR), the system showed BER below 10⁻¹⁰ for storage times up to 50 μs at all temperatures examined except 60 °C, and below 10⁻¹⁰ for storage times < 100 μs at temperatures above 60 °C. The standard acceptable BER in telecommunication systems is cited as 10⁻⁹, and the authors' chosen threshold in this work was 10⁻¹⁰.
• Best combined SNR/BER operating point. Optimal conditions for SNR and BER were found at a cell temperature of 90 °C and a storage time of 5 μs, defined as keeping BER below 10⁻¹⁰ for the longest range of storage times.
• Atomic motion is not the dominant decoherence channel. The authors calculate that going from 353 K to 383 K produces only a 4% increase in average atomic speed, so motional loss should be minimal. They instead point to ground-state population exchange from atomic collisions and increased sensitivity to beam vector fluctuations caused by temperature gradients.
• Noise behaviour at long delays. In noise traces with the probe absent and the control field switched on after 1500 μs, the effect of atomic decoherence between atoms became apparent starting at 1.25 μs.
• Wavelength is not telecom-compatible as built. The system operates at 795 nm (the ⁸⁷Rb D1 line), so photons would require wavelength conversion before deployment in fibre infrastructure; the paper notes rubidium's two-photon resonance permits conversion to telecom wavelengths via a four-wave mixing diamond scheme.

Methodology in Plain English

The memory works by electromagnetically induced transparency. Two stable ground states of ⁸⁷Rb are coupled to a shared excited state by a strong control field and a weak probe field. When both fields are present, interference makes the vapour transparent; when the control field is switched off while the probe is inside the cell, the light is mapped into atomic ground-state populations and slowed to a stop. Switching the control field back on releases the pulse on demand.

Experimentally, a single tunable diode laser (MogLabs CEF795) supplies everything. One tap goes to saturated-absorption locking electronics (Qunnect Qu-Lock), which locks the laser to the ⁵S₁/₂ F = 2 ↔ ⁵P₁/₂ F′ = 1 transition. The probe is derived by an electro-optical modulator driven at the ⁸⁷Rb ground-state splitting, 6834.68 MHz, and the wanted sideband is filtered out with a temperature-controlled etalon of finesse 35. Separate 110 MHz acousto-optic modulators shape both beams in time; the probe pulses are Gaussian with 1.05 μs FWHM and contain roughly 10⁸ photons, and the control field is gated with a step function at 80% duty cycle.

The probe's two polarisation components are split into two spatial rails, rotated, recombined with an orthogonally polarised control field, and converted to circular polarisation to benefit from optical pumping. After the cell, the control field is filtered out with a Glan-Taylor calcite crystal, and a second Sagnac stage recombines the rails into the original input polarisation. Detection is by photodetector and oscilloscope.

For efficiency, the retrieved pulse energy is divided by the input pulse energy after subtracting background taken with the probe blocked (and, for the reference, with the control blocked) and the detector offset. Input traces were averaged over 65536 events and retrieved traces over 64 events to suppress laser amplitude fluctuations. Coherence time comes from an exponential fit to efficiency versus storage time. SNR is the mean over N traces of the time-averaged, offset-corrected signal divided by the RMS of detector fluctuations in the same window. BER is then computed from each SNR value through the Q-function.

Why This Matters

The paper's contribution is less a new physical effect than a bridge: it measures a quantum device with the vocabulary of the telecom industry, which is a practical prerequisite for anyone trying to put quantum memories into deployed fibre plants. Because repeater networks need roughly twice as many memories as there are links, the authors argue that devices must be scalable, economical, low-overhead and near ambient temperature — exactly the properties a warm-vapour cell offers compared with cryogenic or vacuum apparatus.

Real-world applications

• Quantum repeater links and the quantum internet: memory nodes that receive and release photonic qubits on demand to extend entanglement distribution beyond the few tens to few hundred kilometres now reachable by direct fibre transmission.
• Metropolitan quantum key distribution networks: the demonstrated BER below 10⁻¹⁰ for sub-100 μs storage at temperatures above 60 °C places the device within the error budget of standard optical transport.
• Telecom equipment qualification and benchmarking: expressing memory behaviour as SNR in dB and BER gives network planners a common yardstick for deciding where a memory can be inserted into existing architectures.
• Entanglement-distribution infrastructure for distributed quantum computing and sensing: a polarisation-agnostic memory that accepts arbitrary qubit encodings is a reusable building block across several qubit-encoding choices.

Industry relevance: the device studied is a commercial product, the authors credit the vendor (Qunnect) for troubleshooting support, and the metrics chosen are those used to qualify classical optical transceivers. The explicit identification of 795 nm as requiring conversion to telecom wavelengths flags a concrete engineering requirement for product integration.

Future Directions

• Move the measurements into the quantum regime. The authors state that the SNR and BER values here are equipment-dependent and that their instrumentation lacks the sensitivity for quantum measurements. They plan advanced filtering to remove noise from scattering processes and probe field leakage, and to relate SNR to quantum state fidelity.
• Convert SNR and BER into secret key rates. The stated next step is to improve these values for the quantum regime and translate them into key rates rather than classical error metrics.
• Determine entangled photon rates. The authors intend to find entangled photon rates as a function of storage efficiency and pulse width to judge feasibility in real networks.
• Solve the wavelength mismatch. Moving 795 nm photons to telecom wavelengths, for example through the four-wave mixing diamond scheme based on rubidium's two-photon resonance, remains an open engineering task for fibre deployment.
• Reconcile the competing optima. Coherence time is best at 60 °C, efficiency near 70 °C, and SNR/BER at 90 °C, so selecting a single operating point for a deployed node remains an open design question.

Target Audience

Experimentalists and engineers working on quantum memories, quantum repeaters and quantum networking hardware; telecom R&D groups evaluating where quantum devices can enter existing fibre infrastructure; systems integrators assessing commercial quantum components; and graduate students in AMO or quantum information who want a worked example of benchmarking a quantum device against classical communication metrics. Readers looking for a theoretical treatment or a quantum-regime fidelity analysis will find those explicitly deferred to future work.

Authors’ abstract

Quantum repeaters will be paramount to the success of the future quantum internet. Presently however, the losses incurred in fibre communications are unsustainable when fragile quantum states are transmitted over long distances. The most effective repeater nodes will require incorporation of a minimum of two quantum memory units (QMems) for successful operation. Therefore a network composed of numerous nodes will ideally require scalable repeaters that contain QMems that are portable, robust and do not require a high amount of overhead. Here we investigate the performance of a commercial QMem based on a warm vapour of Rb atoms. Specifically, we characterize the coherence time, absolute efficiency, signal to noise ratio and bit error ratio of this device over a range of vapour temperatures that are readily controllable. For our system, we place these performance metrics into the same conventions as those established for current optical communication protocols. Development of QMem systems meeting these benchmarks will be a vital step for seamless integration of QMems systems into present-day communication architectures.