⚛️ quantum physics

A Universal All-Fiber Quantum Buffer for the Telecom Band

This paper presents a room-temperature, fully fiber-integrated quantum buffer operating across the entire telecom C-band that achieves ultra-low loss, long storage times, and high-fidelity preservation of multidimensional qubits, thereby overcoming key integration barriers for scalable quantum networks.

Original authors: Domenico Compagnini, Noemi Tagliavacche, Sara Congia, Andrea Bernardi, Marco Liscidini, Matteo Galli, Daniele Bajoni

Published 2026-06-24

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Domenico Compagnini, Noemi Tagliavacche, Sara Congia, Andrea Bernardi, Marco Liscidini, Matteo Galli, Daniele Bajoni

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the internet as a massive highway system for light. Currently, the "cars" driving on this highway are packets of information. In the future, we want to send "quantum cars" that carry incredibly delicate, secret information. However, these quantum cars are tricky: they often arrive at different times (asynchronously), and to get them to work together, we need to park them temporarily in a garage until the right moment to let them out. This "garage" is called a quantum buffer.

For a long time, building this garage has been a nightmare for engineers. Here is the problem the paper solves, explained simply:

The Old Problem: The "Fragile Freezer" vs. The "Leaky Bucket"

Scientists had two bad options for building these garages:

The "Fragile Freezer" (Matter-based memories): These use special crystals or atoms to hold light. They are great at holding the light for a long time, but they are incredibly picky. They only accept very specific types of light (narrow bandwidth), they need to be kept in super-cold freezers (cryogenics), and they are hard to connect to our existing fiber-optic cables.
The "Leaky Bucket" (Old all-optical buffers): These use mirrors and loops of fiber to hold light. They work at room temperature and accept all kinds of light, but they are very "leaky." Every time the light goes around the loop, a huge chunk of it is lost. After a few turns, the message is gone.

The New Solution: The "Universal Fiber Garage"

The researchers in this paper built a new kind of garage that combines the best of both worlds. It is universal (it works with any type of light), all-fiber (it uses standard cables), and low-loss (it doesn't leak).

Here is how it works, using a creative analogy:

1. The Setup: A Circular Race Track with Smart Gates

Imagine a circular race track made of fiber-optic cable. At the start and finish lines, there are two special gates (called Sagnac interferometers).

Normally, these gates are closed, acting like mirrors that bounce light back and forth, trapping it inside the track.
To let a car (a photon) in, you need to open the first gate.
To let a car out, you need to open the second gate.

2. The Magic Key: The "Control Pulse"

How do you open these gates without breaking the delicate quantum car? You don't use a mechanical key. Instead, you use a "magic key" made of a different color of light (a 1060 nm laser pulse).

When the magic key travels alongside the quantum car inside the fiber, it uses a property called Cross-Phase Modulation. Think of this as the magic key whispering a secret code to the fiber, which instantly changes the "shape" of the gate just for that split second.
This whisper tells the gate: "Open up for the quantum car, but stay closed for everything else."
Once the car is inside, the gate closes again, trapping the car in the loop.

3. The Results: A Perfect Parking Spot

The paper claims this new garage is a massive improvement:

It's almost invisible: When you put a car in and take it out, you lose almost nothing. They measured a loss of only 0.46 dB. In everyday terms, if you put 100 cars in, 90 come out perfectly intact. This is a huge improvement over previous "leaky" systems.
It's a huge garage: The track is long enough to hold the car for about 18 microseconds. While that sounds short, in the world of light, that's an eternity. It allows the system to hold over 200 different cars (temporal modes) at the same time, all parked in a line.
It accepts any car (Universality): The garage doesn't care what kind of car you drive. It can store:
- Time-bin cars: Cars defined by when they arrive.
- Frequency-bin cars: Cars defined by their color (frequency).
- Polarization cars: Cars defined by their orientation (like how a spinning top spins).
- The paper shows it can store all three types perfectly, even if they are entangled (linked) with other cars outside the garage.

4. Why This Matters (According to the Paper)

The authors state that this device solves a major hurdle for the "Quantum Internet."

Room Temperature: You don't need a freezer. It works right on a desk.
Standard Cables: It uses the same cables already buried underground for the internet.
Synchronization: It allows different quantum signals, which arrive at random times, to be lined up perfectly so they can talk to each other.

In summary: The paper presents a "universal parking garage" for light. It uses a clever trick with a secondary laser to open and close doors in a fiber-optic loop. It keeps the light safe with almost no loss, works at room temperature, and can handle many different types of quantum information at once, making it a ready-to-use tool for building future quantum networks.

Technical Summary: A Universal All-Fiber Quantum Buffer for the Telecom Band

Problem Statement
The realization of a scalable quantum internet requires the ability to temporally align asynchronous photonic signals through on-demand buffering. Existing quantum memory technologies face a fundamental trade-off between performance metrics and practical integration. Matter-based memories (e.g., atomic vapors, rare-earth-ion-doped crystals) offer long storage lifetimes but suffer from narrow operational bandwidths (MHz to GHz regime) and strict cryogenic requirements, hindering integration with standard telecommunications infrastructure. Conversely, existing all-optical memories and buffers operate at room temperature with broad bandwidths but are typically limited by high input/output losses (several decibels per cycle) and a lack of universality across different photonic degrees of freedom (DoF). Consequently, a single quantum memory capable of simultaneously supporting time-bin, frequency-bin, and polarization encodings at telecommunications wavelengths with low loss has remained elusive.

Methodology
The authors present a fully fiber-integrated quantum buffer operating in the telecommunications C-band (1550 nm). The device architecture consists of two Sagnac interferometers constructed from standard SMF-28 Ultra optical fiber (100 m each), connected to form a cavity. The system utilizes a "loop-and-switch" mechanism driven by the ultrafast Kerr nonlinearity ( $\chi^{(3)}$ ) of the optical fiber.

Switching Mechanism: The Sagnac interferometers function as dynamically switchable mirrors. An intense optical control pulse at 1060 nm is injected to overlap with the signal pulse within the interferometer. Through Cross-Phase Modulation (XPM), the control pulse imparts a $\pi$ phase shift exclusively to the co-propagating component of the quantum signal. This phase shift alters the interference condition, switching the Sagnac loop from a reflective state (trapping the photon) to a transmissive state (injecting or ejecting the photon).
Wavelength Management: The beamsplitters closing the Sagnac loops are wavelength-dependent, acting as 50:50 splitters for the 1550 nm signal but as transparent pass-throughs for the 1060 nm control pulses. This design eliminates the need for intra-loop wavelength-division multiplexers or active electro-optic switches, which typically introduce significant insertion losses.
Operation: The buffer operates at room temperature. The round-trip time is determined by the fiber length (approx. 100 m loops + 5 m connection), set to $1.03 \, \mu\text{s}$ . Storage is achieved by trapping the photon between the two Sagnac loops, and retrieval is triggered by a second control pulse timed to overlap with the circulating photon in the second interferometer.

Key Contributions and Results

Ultra-Low Loss and High Efficiency:
The device achieves a record-low end-to-end input/output insertion loss of 0.46 dB (90% transmission). The internal round-trip loss is measured at 0.24 dB, enabling a characteristic storage lifetime of $\sim 18 \, \mu\text{s}$ (corresponding to $\sim 18$ round-trips).
Broadband Operation and Time-Bandwidth Product (TBP):
The buffer exhibits an operational bandwidth exceeding 12.5 THz ( $\sim 100$ nm), covering the full telecom C-band. This results in an exceptionally large Time-Bandwidth Product of $2.2 \times 10^8$ , outperforming most state-of-the-art matter-based systems by several orders of magnitude.
Universal Compatibility (Multi-DoF):
The platform demonstrates high-fidelity storage for all three degrees of freedom compatible with standard single-mode fiber:
- Time-bin qubits: Fidelities exceeding 99% (self-referenced) across all storage loops.
- Polarization qubits: Fidelities exceeding 99% with active compensation for residual birefringence.
- Frequency-bin qubits: Fidelities exceeding 99% for adjacent frequency components separated by 1 GHz.
  The device preserves the spectral and temporal profiles of the stored photons without introducing measurable decoherence or distortion.
Entanglement Preservation:
The authors demonstrate the robust preservation of quantum entanglement. In a memory-node configuration where one photon of an entangled pair is stored while its partner propagates, the retrieved photons maintain high-quality non-local correlations. Quantum interference visibilities reach up to 98.8%, violating the CHSH inequality by more than 20 standard deviations, with no observable degradation as a function of storage time.
High-Capacity Temporal Multiplexing:
The buffer supports the simultaneous storage of over 200 independent temporal modes. The authors demonstrate the ability to address these modes either collectively or individually. By interleaving control pulses, they successfully stored and retrieved packets of 20 modes, achieving a total circulating capacity of 200 modes within the compact architecture.

Significance
The paper claims that this work establishes a "truly universal quantum buffer" that resolves the long-standing dichotomy between low-loss storage and massive optical bandwidth. By operating at room temperature with ultra-low loss and full compatibility with the telecom C-band, the device provides a robust solution for the high-rate synchronization of multidimensional quantum states.

The authors position this technology not as a long-term storage medium (like matter-based memories with millisecond-to-hour lifetimes) but as a high-performance quantum cache or buffer. Its significance lies in enabling:

Scalable Quantum Networking: Facilitating the synchronization of probabilistic photon sources and the temporal reordering of quantum information within network nodes.
Multiplexing: Enabling high-capacity time-domain multiplexing for intranode architectures, which is critical for overcoming the exponential scaling bottlenecks of multi-photon generation probabilities.
Integration: Offering a plug-and-play solution for existing fiber-optic infrastructure without the need for cryogenics, free-space optics, or complex frequency conversion.

The results confirm that the all-optical approach can circumvent the traditional trade-offs between bandwidth, storage time, and efficiency, clearing a major hurdle for the deployment of global photonic quantum networks.