Linear-optical test of quantum contextuality with sequential measurements

This paper presents and experimentally demonstrates a robust linear-optical setup using sequential measurements with single photons to violate the KCBS inequality, thereby verifying Kochen-Specker contextuality and providing a practical tool for characterizing single-photon sources.

The Gist

The Big Idea: Why the "Context" Matters

Imagine you are at a restaurant. You order a burger.

• Scenario A: You eat the burger with a side of fries.
• Scenario B: You eat the burger with a side of salad.

In the classical world (the world of everyday objects), the burger tastes exactly the same regardless of what you eat with it. Its "flavor" is an intrinsic property of the burger itself.

In the quantum world (the world of tiny particles like photons), this isn't true. This paper is about a phenomenon called Quantum Contextuality. It proves that for quantum particles, the "flavor" of a measurement depends entirely on what else you are measuring at the same time. The result changes based on the "context" (the company the measurement keeps).

If the universe worked like a classical restaurant, the burger's taste would be fixed. But quantum mechanics says the universe is more like a magical menu where the taste of the burger changes depending on whether it's paired with fries or salad.

The Problem: The "Destructive" Camera

To prove this, scientists usually need to measure a particle, then measure it again immediately after to see if the context changed the result.

Here is the catch: In the world of light (photons), measuring a particle is usually like taking a photo with a flash that destroys the subject. Once you "click" the detector to see the photon, the photon is gone. You can't measure it a second time.

Previous experiments tried to get around this by using clever tricks, but they had a flaw: they didn't measure the exact same thing twice. It was like measuring the burger in Scenario A, then swapping it for a slightly different burger to measure in Scenario B. That doesn't prove the context changed the taste; it just proves the burgers were different.

The Solution: The "Ghost" Detector

The authors of this paper built a new machine using linear optics (mirrors, beam splitters, and lenses) and a special type of detector that acts like a "ghost."

Here is how their trick works:

1. The Setup: They send a single photon through a maze of mirrors.
2. The "Click" vs. "No-Click": They use a detector that can either "click" (saying "I see a photon!") or stay silent ("No-click").
3. The Magic: If the detector clicks, the photon is absorbed and destroyed (game over). But if the detector stays silent (a "no-click"), the photon wasn't there. Because the photon wasn't in that specific spot, it wasn't destroyed. It continues traveling through the rest of the maze to be measured again.

Think of it like a security guard at a door.

• If the guard sees you (click), you are stopped and removed.
• If the guard doesn't see you (no-click), you are allowed to walk through the door and keep going.

By only looking at the times the guard didn't see the photon, the scientists can measure the photon, let it pass through, and measure it again. This allows them to perform a sequential measurement without destroying the particle.

The Experiment: The KCBS Inequality

The team used a famous mathematical rule called the KCBS inequality.

• The Rule: If the universe works like a classical restaurant (where the burger has a fixed taste), a specific math formula involving five different measurements must always add up to a number greater than -3.
• The Result: When the scientists ran their experiment with single photons, the number came out to roughly -3.94.

Because -3.94 is lower than -3, the "classical rule" was broken. This proves that the photon's behavior did depend on the context of the measurement. The "burger" really did taste different depending on its neighbors.

Why This Matters (According to the Paper)

1. It's a True Test: Unlike previous experiments, this setup ensures that the exact same physical operation is used every time a measurement is made, just in a different order. This closes a loophole that critics had pointed out before.
2. It's Tough: The experiment still worked even when they simulated "photon loss" (like a photon getting lost in the maze). It remained valid even if about 10% of the photons were lost.
3. It's a Tool: Beyond proving quantum mechanics is weird, the authors say this setup can be used as a practical tool. If you have a light source and want to know if it's truly a "single-photon source" (a machine that spits out exactly one photon at a time), you can run this test. If the math works out, you know you have a high-quality single photon. If it fails, your light source might be leaking extra photons or vacuum (empty space).

Summary

The paper describes a clever way to measure a single photon twice in a row without destroying it, by using a "silent" detector that lets the photon pass if it's not there. Using this method, they proved that quantum particles change their behavior based on what else is being measured around them, violating a classical rule of physics. They also showed this method is robust and can be used to verify the quality of single-photon light sources.

Technical Summary

Technical Summary: Linear-optical test of quantum contextuality with sequential measurements

Problem Statement
Quantum contextuality, the dependence of measurement outcomes on the set of jointly measured compatible observables, serves as a fundamental signature of nonclassical behavior that cannot be explained by noncontextual hidden-variable models. While the Kochen–Specker (KS) theorem and associated inequalities, such as the Klyachko–Can–Binicioğlu–Shumovsky (KCBS) inequality, provide theoretical frameworks for testing this phenomenon, experimental verification in photonic systems faces significant challenges.

The primary obstacle is the destructive nature of photodetection, which complicates the implementation of sequential measurements required to evaluate joint observables accurately. Previous experimental approaches have attempted to circumvent this by encoding measurement outcomes into different degrees of freedom (e.g., polarization or path) to perform joint measurements rather than true sequences. However, as noted in the literature, such strategies do not fully capture the quantum nature of contextuality and yield data that can be classically simulated. Furthermore, other implementations utilizing sequential detectors often fail to ensure that the same observable is realized by identical physical operations across different contexts (e.g., measuring $A^{(5)}A^{(1)}$ versus $A^{(5)}A'^{(1)}$ where $A^{(1)}$ and $A'^{(1)}$ are physically distinct). This paper addresses the need for a linear-optical scheme that realizes genuine sequential measurements, ensures the physical identity of observables across contexts, and overcomes the destructive nature of detection.

Methodology
The authors propose and experimentally implement a linear-optical setup using single photons distributed over $M \ge 3$ bosonic modes. The core of the methodology involves:

1. Sequential Measurement via No-Click Events: Instead of using non-demolition detectors (which are resource-intensive) or joint measurements, the scheme employs on–off photodetectors. A "click" (detection of a photon) corresponds to the outcome $-1$ and destroys the photon. Crucially, a "no-click" event (vacuum detection) corresponds to the outcome $+1$. This no-click event projects the system onto the subspace of the remaining modes without destroying the single photon, thereby allowing a subsequent measurement to be performed on the same photon. This satisfies the repeatability and non-disturbance requirements essential for contextuality tests.
2. Physical Identity of Observables: The setup utilizes linear-optical networks (unitary transformations $U_j$) sandwiched around a fixed on–off detector. By defining the observable $A^{(j)} = U_j D U_j^\dagger$, where $D$ is the fixed detector observable, the scheme ensures that whenever an observable appears in a correlation (e.g., $A^{(1)}$ in the pair $A^{(1)}A^{(2)}$ and $A^{(5)}A^{(1)}$), it is implemented by the exact same physical operation. This resolves the issue of non-identical compatible measurements found in prior experiments.
3. KCBS Inequality Formulation: The experiment tests a simplified form of the KCBS inequality expressed in terms of joint probabilities:
   $$S(\rho) = \sum_{j=1}^5 p(v_j = +1) - \sum_{j=1}^5 p(v_j = +1, v_{j+1} = +1) \le 2$$
   where the classical bound is 2. In the proposed scheme, the term $p(v_j = +1, v_{j+1} = +1)$ corresponds to the probability of two consecutive "no-click" events.
4. Experimental Implementation: The experiment uses a hybrid encoding of spatial paths and polarization to realize three optical modes ($M=3$). A heralded single-photon source generates photons, which are routed through a sequence of beam displacers (BDs) and half-wave plates (HWPs) to implement the required unitary transformations. The setup includes a mechanism to simulate photon loss by coupling a fraction of the light to a vacuum state, allowing for the study of robustness against loss.

Key Contributions

• Novel Sequential Scheme: The paper introduces a linear-optical architecture that realizes true sequential measurements using standard on–off detectors by leveraging vacuum (no-click) events. This avoids the need for complex quantum non-demolition interactions.
• Resolution of Contextuality Implementation Issues: The construction guarantees that each co-measured observable is implemented by the same physical operation across different contexts, addressing a long-standing limitation in photonic contextuality tests.
• Application to State Characterization: The authors demonstrate that the setup can be used beyond fundamental tests to verify nonclassicality and probe the photon-number statistics of arbitrary single-mode states. They derive bounds showing that observing $S(\rho) > 1.25$ certifies nonclassicality, and specific thresholds (e.g., $S > 1.472$) can certify the presence of a single-photon component, even in the presence of vacuum or higher-order photon-number components where standard $g^{(2)}$ tests might fail.

Experimental Results
The experimental results demonstrate a clear violation of the KCBS inequality.

• Violation Magnitude: For an initial single-photon state, the measured value is $S(|1\rangle_1) = 2.2363 \pm 0.0041$, which exceeds the classical bound of 2 and is close to the theoretical maximum quantum violation of $\sqrt{5} \approx 2.236$.
• Robustness to Loss: The experiment investigated the effect of photon loss by varying the ratio of single-photon to vacuum components. The violation of the inequality ($S > 2$) was maintained up to a photon loss rate of approximately 10.55%. The data confirms that the observed value of $S$ decreases linearly with the single-photon fraction, consistent with theoretical predictions.
• Statistical Significance: The results were averaged over 100 repetitions with accumulation times of 1 second per cycle, yielding statistical uncertainties that confirm the violation is significant.

Significance and Claims
The paper claims that this work provides a practical and robust tool for probing nonclassicality and verifying single-photon sources. By resolving the issue of identical observable realization and utilizing a simple linear-optical setup, the scheme offers a feasible route for operational or "black-box" verifications of contextuality. The authors assert that their approach simplifies experimental implementation while fulfilling the core Kochen–Specker requirements. Furthermore, the ability to infer photon-number statistics and certify properties of input states suggests applications in the self-testing and certification of photonic components, including single-photon sources and detectors. The results provide strong evidence against noncontextual hidden-variable descriptions of the observed correlations and highlight the interplay between the wave-like (superposition) and particle-like (mutual exclusivity of clicks) aspects of single photons.