⚛️ quantum physics

Testing Single Photon Entanglement using Self-Referential Measurements

This paper demonstrates a violation of Bell inequalities using self-referential joint measurements on two copies of a single-photon entangled state, offering a more accessible experimental route to verify single-photon nonlocality without the complexity of homodyne measurements.

Original authors: Daniel Kun, Teodor Strömberg, Borivoje Dakić, Philip Walther, Lee A. Rozema

Published 2026-04-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Daniel Kun, Teodor Strömberg, Borivoje Dakić, Philip Walther, Lee A. Rozema

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

The Big Idea: Proving "Spooky Action" with Just One Photon

For decades, physicists have been fascinated by quantum entanglement. Usually, we think of this as a pair of twins who are magically connected: if you check one twin in New York and find them wearing a red hat, you instantly know the other twin in London is wearing a blue hat, even though they haven't spoken.

But what if you only have one photon (a particle of light)? Can a single particle be "entangled"?

Thirty years ago, scientists predicted that if you send a single photon through a splitter (like a fork in the road), it travels down both paths at once. One path goes to "Alice," the other to "Bob." In a way, the photon is entangled with itself across two locations. However, proving this is tricky because it requires very complex, messy equipment (called homodyne measurements) that acts like a shared ruler between Alice and Bob. Critics argued that this "shared ruler" might be cheating the test.

This paper presents a clever new trick: Instead of using one photon and a complex ruler, the team used two identical copies of that single-photon setup. They let the two copies act as "rulers" for each other. This allowed them to prove the "spooky connection" exists without the messy equipment, using a method they call self-referential measurement.

The Analogy: The "Twin Coin" Game

Imagine a game played by Alice and Bob, who are in separate rooms.

The Old Way (The Problem)

In the old experiments, to prove they were connected, Alice and Bob needed a Master Clock that was split in half and sent to both of them. They had to use this clock to synchronize their measurements.

The Flaw: Critics said, "Wait a minute! If you split a laser beam to make that clock, the two halves might already be secretly talking to each other before the game even starts. So, maybe the 'spooky connection' isn't real; it's just the clock cheating!"

The New Way (The Solution in this Paper)

The team in this paper said, "Let's try a different game. We won't use a Master Clock. Instead, we will give two identical sets of coins to Alice and Bob."

The Setup:
- We create Photon 1. It gets split, so part of it goes to Alice and part to Bob.
- We create Photon 2 (an identical twin). It also gets split, with parts going to Alice and Bob.
- Crucially: Photon 1 and Photon 2 are strangers. They don't know each other yet.
The Measurement (The "Self-Reference"):
- Alice now has two pieces of light (one from Photon 1, one from Photon 2). Bob has the other two.
- Instead of needing an outside clock, Alice uses Photon 2 to measure Photon 1.
- Think of it like this: If you want to know if a coin is heads or tails, you don't need a ruler. You just flip a second, identical coin next to it and compare them.
- In this experiment, the two photons act as mirrors for each other. They "reference" each other to see if their patterns match.
The Result:
- When Alice and Bob compare their results, they find a correlation that is impossible to explain with normal physics.
- The numbers they got (called the CHSH parameter) were 2.71 and 2.23.
- In the world of local physics (where things can't talk faster than light), the maximum score you can get is 2.
- Since they got higher than 2, they proved that the single photon really was in two places at once, and the "spooky connection" is real.

Why This Matters

Simplicity: The old method required complex lasers and "local oscillators" (the Master Clocks) that were hard to build and easy to misunderstand. This new method is like swapping a supercomputer for a simple pair of dice. It's much easier to build and understand.
No Cheating: Because the two photons start as strangers and only interact when Alice and Bob measure them, there is no way for them to have "pre-arranged" a secret code. This closes a loophole that critics had used to doubt previous experiments.
Future Tech: This technique could be used for other types of particles (not just light) and could help build better quantum computers and sensors.

The Bottom Line

The researchers took a difficult, controversial problem (proving a single particle can be in two places at once) and solved it with a clever workaround. By using two copies of the experiment to measure each other, they created a "self-referential" system that proved quantum entanglement is real, without needing the complex, controversial equipment of the past.

It's like proving two people are twins by having them compare notes with each other, rather than trying to find a birth certificate that might have been forged. The result? The twins are definitely connected.

1. Problem Statement

Quantum entanglement is typically demonstrated using pairs of particles shared between two parties (Alice and Bob). However, theoretical predictions dating back thirty years suggest that single-photon entanglement (mode-entanglement) can also violate Bell inequalities. In this scenario, a single photon is sent through a beam splitter, placing it in a superposition of traveling to Alice or Bob ( $|\psi\rangle = \frac{1}{\sqrt{2}}(|0,1\rangle + |1,0\rangle)$ ).

Previous experimental demonstrations of single-photon nonlocality relied on homodyne measurements. These methods require Alice and Bob to share a common phase reference (a Local Oscillator or LO), typically created by splitting a laser beam. This approach faces two significant challenges:

Technical Complexity: Homodyne detection is technically demanding, especially for mode-entangled states involving massive particles.
Interpretational Loophole: There is a long-standing debate regarding whether the shared LO itself carries nonlocal correlations after being split. If the LO is not a separable state, the observed Bell violation might be attributed to the reference beam rather than the single-photon state, casting doubt on the conclusiveness of the test.

2. Methodology

The authors propose and experimentally realize a self-referential measurement scheme that eliminates the need for an external shared phase reference.

State Preparation: Instead of a single photon, the experiment utilizes two independent but identical copies of the single-photon entangled state.
- Two indistinguishable photons ( $\gamma_1, \gamma_2$ ) are generated via Spontaneous Parametric Down-Conversion (SPDC).
- Each photon is sent to a separate 50/50 beam splitter, creating two copies of the state: $|\psi_{in}\rangle = \frac{1}{2}(|0,1\rangle_{A1B1} + |1,0\rangle_{A1B1}) \otimes (|0,1\rangle_{A2B2} + |1,0\rangle_{A2B2})$ .
- Modes $A_1, A_2$ are sent to Alice; modes $B_1, B_2$ are sent to Bob. Crucially, there is no initial entanglement between photon 1 and photon 2.
Self-Referential Mechanism:
- Alice and Bob perform joint measurements on their respective pairs of modes ( $A_1, A_2$ and $B_1, B_2$ ).
- They locally interfere these modes on a beam splitter.
- In this setup, photon 2 acts as a phase reference for photon 1, and vice versa. This internal referencing removes the need for an external LO, making the measurement "self-referential."
Detection Strategy:
- The setup uses pseudo-number-resolving detectors (fiber beam splitters followed by two single-photon detectors per port) to distinguish between single clicks and double clicks.
- Post-selection vs. Non-post-selection:
  - With Post-selection: Events where both photons hit the same detector (double clicks) are discarded.
  - Without Post-selection: All detection patterns, including double clicks and in-lab coincidences (Hong-Ou-Mandel dips), are included in the analysis.
Bell Test: Alice and Bob choose between two phase settings ( $\phi_x, \phi_y$ ) to define their measurement bases. They calculate the CHSH parameter ( $S$ ) based on the correlation of their joint outcomes.

3. Key Contributions

Novel Protocol: The paper introduces a method to violate a Bell inequality using single-photon entanglement without a shared external phase reference, thereby closing the "shared reference" loophole associated with homodyne schemes.
Theoretical Derivation: The authors derive the specific CHSH bounds for this two-copy scenario. They show that:
- With post-selection, the maximum theoretical violation reaches the Tsirelson bound ( $2\sqrt{2} \approx 2.828$ ).
- Without post-selection (including all events), the maximum theoretical violation is $1 + \sqrt{2} \approx 2.414$ .
Experimental Realization: They successfully implemented this protocol using standard photonic components, demonstrating that single-photon nonlocality can be verified with relatively simple joint measurements rather than complex homodyne setups.

4. Experimental Results

The experiment was conducted using indistinguishable photons generated via SPDC, with high spectral and polarization filtering.

Indistinguishability: The Hong-Ou-Mandel (HOM) visibility was measured to be 95.2%, confirming the high indistinguishability of the photon pairs.
Bell Correlation Visibility: The measured visibility of the Bell correlations was 91.4% (with post-selection) and 90.2% (without post-selection).
CHSH Parameter Violations:
- With Post-selection: The experiment achieved a CHSH parameter of $\tilde{B} = 2.71 \pm 0.09$ . This exceeds the classical limit of 2 and approaches the theoretical maximum of $2\sqrt{2}$ .
- Without Post-selection: The experiment achieved a CHSH parameter of $B = 2.23 \pm 0.07$ . This exceeds the classical limit of 2 and approaches the theoretical maximum of $1+\sqrt{2}$ .
Robustness: The results hold even when including "double click" events, which contribute classical correlations, proving that the nonlocality is intrinsic to the single-photon state structure rather than an artifact of post-selection.

5. Significance

Foundational Clarity: This work provides a conclusive test of single-photon nonlocality. By removing the shared LO, it definitively demonstrates that the violation arises from the entanglement of the single photon's modes, not from correlations in a reference beam.
Accessibility: The method avoids the technical complexity of homodyne detection, offering a more accessible experimental route for testing nonlocality.
Scalability: The technique is applicable to general mode-entangled states, including those involving massive particles (where homodyne detection is difficult) and higher-order states like N00N states.
Technological Impact: This approach could simplify the implementation of quantum communication and sensing protocols that rely on mode-entanglement, as it removes the overhead of distributing stable phase references between distant nodes.