⚛️ quantum physics

Looking Beyond Visibility: Resolving the Temporal Structure of Continuous-Wave Four-Photon Hong-Ou-Mandel Interference

This paper presents a method to resolve the temporal structure of continuous-wave four-photon Hong-Ou-Mandel interference by recording all photon arrival times to create a three-dimensional coincidence histogram, thereby preserving critical correlation details and enabling post-processed optimization of coincidence windows for quantum network protocols.

Original authors: Sabine Häussler, Meritxell Cabrejo-Ponce, Ana Maria Torres, Baghdasar Baghdasaryan, Karen Lozano-Mendéz, Fabian Steinlechner

Published 2026-10-16

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Sabine Häussler, Meritxell Cabrejo-Ponce, Ana Maria Torres, Baghdasar Baghdasaryan, Karen Lozano-Mendéz, Fabian Steinlechner

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

In the world of quantum physics, light is not just a beam that illuminates a room; it is a stream of individual particles called photons that can be linked together in mysterious ways. When two such photons meet at a special mirror called a beam splitter, they can interfere with one another, behaving in a way that reveals whether they are identical twins or strangers. This phenomenon, known as Hong-Ou-Mandel interference, is the foundation for many future technologies, including ultra-secure communication networks and powerful quantum computers. To make these networks work, scientists often use light sources that emit photons continuously, like a steady stream of water from a hose, rather than in short, synchronized bursts. However, working with this continuous flow creates a unique challenge: because the photons arrive at random times, it is difficult to tell which ones are actually matching pairs and which are just random noise. Without a way to distinguish the true pairs, the delicate quantum connections needed for these technologies cannot be formed reliably.

A team of researchers in Germany has developed a new way to solve this problem by looking at the timing of photons with extreme precision. Instead of trying to force the photons to arrive at the same time using complex hardware, they recorded the exact moment every single photon arrived at their detectors. They used two tiny, ring-shaped devices made of silicon nitride to generate pairs of photons. These rings act like resonators, trapping light and releasing it in pairs that are linked by their properties. The researchers set up an experiment where photons from two separate rings were sent to a beam splitter to see if they would interfere. Crucially, they did not just count how many times four photons were detected together; they captured the precise arrival time of each one relative to a starting point. This created a massive, three-dimensional map of the data, showing how the photons were related in time.

By analyzing this map, the team discovered that the interference pattern was not a simple line but a complex structure with distinct branches. One branch represented cases where both photons passed through the beam splitter, while another showed cases where both were reflected. In the center, where these branches met, the number of detections dropped sharply, creating a "dip" that proved the photons were interfering. The researchers found that by adjusting how they defined a "coincidence"—essentially deciding how close in time two photons must be to count as a pair—they could trade off between the number of successful events and the clarity of the interference. If they set the time window too wide, they included too much random noise, which blurred the results. If they set it too narrow, they lost too many valid events. However, by using their new method of looking at the full three-dimensional timing data, they could optimize this window to get the best possible result for their specific needs.

The study showed that with a time window of 600 picoseconds (a picosecond is one trillionth of a second), they could achieve a visibility of about 85 percent, meaning the interference was very clear, while still getting hundreds of events per hour. When they tightened the window to 300 picoseconds, the clarity improved to over 93 percent, but the number of events dropped significantly. This trade-off is vital because different applications require different balances between speed and precision. The researchers also built a computer model that accurately predicted these results, showing that their understanding of the physics was correct. This model allows them to estimate the performance of their system before even running the full experiment, using only basic measurements of the light sources.

This work is significant because it provides a clear, standardized way to measure and improve continuous-wave quantum sources. Previous experiments often used vague or inconsistent methods to filter their data, making it hard to compare results from different labs. The new approach offers a general benchmark that can be applied to any system using continuous light. It proves that by using advanced time-resolved detection and smart post-processing, scientists can extract high-quality quantum information from continuous sources without needing complex synchronization hardware. This paves the way for more compact and efficient quantum networks, where independent light sources can be linked together to share information securely across long distances. The ability to tune the balance between the quantity of data and the quality of the quantum connection gives engineers a powerful new tool for building the quantum internet of the future.

Technical Summary: Resolving the Temporal Structure of Continuous-Wave Four-Photon Hong-Ou-Mandel Interference

Problem Statement
Hong-Ou-Mandel (HOM) interference is a foundational element for photonic quantum information protocols, including entanglement certification and quantum key distribution. While traditionally characterized by a single visibility parameter obtained by averaging over a fixed coincidence window, this approach discards critical details of the underlying temporal structure, particularly in continuous-wave (CW) processes. In CW systems, where synchronization is often shifted to the electronic domain via post-processing, the lack of standardized temporal postselection strategies makes cross-experimental comparison difficult. Furthermore, four-photon interference with CW sources requires photon pairs with long coherence times, necessitating a method to resolve the complex temporal correlations without relying on physical optical delay lines.

Methodology
The authors present a systematic experimental and post-processing framework for analyzing asynchronous four-fold HOM interference between independent CW sources. The methodology comprises three main components:

Experimental Setup: The experiment utilizes two independent silicon nitride microring resonators (MRRs) pumped by a 1550 nm CW laser to generate correlated photon pairs via spontaneous four-wave mixing (SFWM). The signal photons serve as heralds (detectors A and D), while the idler photons interfere at a 50:50 beam splitter (detectors B and C). The system employs superconducting nanowire single-photon detectors (SNSPDs) with ultrafast time-tagging capabilities (1 ps resolution), allowing for the recording of all four photon arrival times without physical delay lines.
3D Temporal Mapping: Instead of projecting data into a 1D histogram immediately, the authors map four-fold events into a three-dimensional (3D) delay space defined by relative delays $\tau_{AB} = t_B - t_A$ , $\tau_{AC} = t_C - t_A$ , and $\tau_{AD} = t_D - t_A$ . This representation preserves relative arrival-time correlations that are typically discarded.
Theoretical Modeling and Post-Processing: A simulation model is developed to describe the four-fold coincidence rates, accounting for:
- True Interference: The destructive interference between reflected-reflected (RR) and transmitted-transmitted (TT) paths, modeled using double-exponential decay functions derived from the Lorentzian spectral profile of the MRRs.
- Accidental Coincidences: A detailed classification of accidental four-fold events (e.g., one correlated pair plus two uncorrelated detections), which manifest as distinct planes in the 3D delay space.
- Extraction Strategies: Two methods for extracting 1D HOM curves from the 3D data are compared:
  - Method 1 (Single-Channel Delay): Scanning the delay of one heralding channel while keeping others fixed. This requires wide coincidence windows to establish a baseline, often reducing visibility due to accidental counts.
  - Method 2 (Dual-Channel Delay): Scanning the delays of both channels belonging to the same photon pair simultaneously. This preserves the temporal correlation within pairs, allowing for narrow coincidence windows and accurate baseline determination ( $R_\infty$ ) even with high temporal filtering.

Key Results

3D Structure Visualization: The experimental 3D histogram clearly reveals two diagonal branches corresponding to RR and TT events, intersecting at the origin where HOM interference suppresses coincidences. Accidental events are observed as four distinct planes crossing these branches.
Method Comparison: Method 2 is shown to be superior for CW systems. It enables the reconstruction of a well-defined HOM dip with narrow coincidence windows (e.g., 300 ps), whereas Method 1 fails to produce a usable baseline with such narrow windows.
Visibility vs. Count Rate Trade-off: The study demonstrates a tunable trade-off between interference visibility and count rate. By adjusting the coincidence window size, the authors achieved a visibility of 93.2% with a count rate of 83 counts/hour (at 300 ps window) and 84.6% visibility with 387 counts/hour (at 600 ps window).
Model Accuracy: The theoretical model, which requires only single-count rates and two-fold coincidence histograms as input, accurately predicts the behavior of the four-fold count rates and the evolution of the HOM dip. The model serves as an upper bound for performance, with deviations attributed to experimental imperfections such as partial polarization distinguishability and spectral drift.
Accidental Correction: The authors quantify the contribution of accidental coincidences, showing that correcting for them reveals the intrinsic source quality (e.g., increasing measured visibility from 93.2% to 96.8% after correction), though they note this correction should not be applied when modeling quantum key distribution performance where background noise is physically present.

Significance and Claims
The paper claims to provide a general benchmark for continuous-wave sources in interference-based quantum network protocols. Its primary significance lies in shifting the characterization of HOM interference from a single scalar value (visibility) to a resolved temporal structure. The authors emphasize that the different reconstruction methods (Method 1 vs. Method 2) are not merely numerical variations but represent physically distinct temporal post-selection strategies that yield different, yet equally valid, visibility values. Consequently, the paper argues that any reported visibility must be accompanied by a clear definition of the post-selection procedure, the coincidence windows used, and the corresponding count rate. This approach facilitates the optimization of quantum protocols by allowing researchers to trade off visibility against count rate based on specific application requirements, leveraging the asynchronous nature of CW sources to avoid the need for stringent optical synchronization.

More like this