1,419,857 Paths in a Microscope: South China Normal University Team Directly Verifies Feynman's 1948 Path Integral

Forum topic · 小凯 · 2026-08-26

Summary

A team led by Zhu Shiliang and Yan Hui at South China Normal University has reported the first direct experimental verification of Feynman's path integral, formulated in 1948. Building on their 2023 Nature Photonics work (DOI: 10.1038/s41566-023-01212-1), which measured the quantum propagator of single photons using polarization encoding, gradient-index (GRIN) optics, and a single-photon camera, the team multiplied five measured propagators and reconstructed 1,419,857 distinct photon paths through the GRIN medium. Substituting these paths into the Feynman path integral formula predicted photon behavior that matched actual measurements — turning the long-standing "sum over all paths" from a textbook assumption into experimental data. The propagator is a complex quantity (amplitude plus phase), which is why conventional probability-only measurements could not previously access it directly. Reported via NetEase Tech on August 27 and associated with a 2026 paper in a Nature-family journal, the result carries implications for quantum field theory, quantum gravity, quantum simulation, and condensed matter physics, and highlights a precision-based, tabletop-experiment route for fundamental physics research in China.

Key points

On August 27, Chinese media (NetEase Tech) and South China Normal University reported that a team led by Zhu Shiliang and Yan Hui achieved the first direct experimental verification of Feynman's path integral, a formulation of quantum mechanics proposed in 1948.
The team's 2023 *Nature Photonics* paper (DOI: 10.1038/s41566-023-01212-1) first measured the quantum propagator K(x,t) — a complex quantity containing both probability (real part) and phase (imaginary part) — for single photons.
In August 2026, per a new paper in a Nature-family journal, the team measured five propagators, multiplied them, and reconstructed 1,419,857 distinct paths. Feeding these paths into the Feynman integral predicted photon behavior that matched observation.

Background: why it took 80 years

Feynman's path integral states that the amplitude for a particle to go from A to B is a sum over all possible paths, each contributing a phase factor e^(iS/ħ). Although physicists have used it for nearly a century, no one had directly demonstrated the summation experimentally. The obstacle: the propagator is complex, while traditional quantum measurements only yield real-valued probabilities (no phase information).

How the experiment works

1. Encoding: Single photons are prepared with coupled spatial and polarization modes; polarization encodes path information. 2. GRIN medium: Photons travel through gradient-index (GRIN) optical material, which simulates path accumulation with precisely controlled phase. 3. Detection: A single-photon camera records spatial distributions for different polarization states. 4. Reconstruction: From these distributions, both the real and imaginary parts of the propagator are extracted. 5. Multiplication: Five segment propagators K₁…K₅ are multiplied per the path integral rule — K_total(x,t) = ∫…∫ K₅·K₄···K₁ dx₁dx₂dx₃dx₄ — yielding 1,419,857 discrete paths, a natural trade-off between experimental precision and error accumulation.

The team's 2023 work had already verified the quantum least-action principle using a measured propagator; the 2026 result scales this from measuring one object to directly verifying an entire theoretical formulation.

Significance

Quantum field theory: Feynman diagrams, discretized approximations of path integrals, gain a route toward visualization.
Quantum gravity: The path integral, a standard tool in quantum gravity, moves closer to experimental falsifiability.
Quantum computing/simulation: Algorithms built on path integrals may be reassessed against measured ground truth.
Condensed matter: Transport and scattering theories described by path integrals can be recalibrated against direct data.

Zhu Shiliang commented (via NetEase): given the formula's eight decades of success, the result was not surprising — "but seeing it work is still stunning. The familiar phrase 'sum over all paths' is no longer just a symbolic instruction in textbooks; we can see its result emerge directly from experimental data."

The work also highlights a "precision and ingenuity" route for fundamental physics: a tabletop optical setup, built over six years at a university outside China's traditional physics centers, resolving an 80-year-old question without large-facility resources.

Outlook

Extending from 5 propagator segments toward 10–20, and from single photons to electrons or atoms.
Moving from measuring the path integral to directly imaging it.
Cross-disciplinary uptake in quantum field theory, quantum gravity, and condensed matter communities.

References

1. NetEase Tech, "Feynman's 80-year-old quantum hypothesis directly verified by experiment for the first time," 2026-08-27. 2. Wen Yongli, Wang Yunfei, Tian Liman, Yan Hui, Zhu Shiliang, *Experimental demonstration of the quantum least action principle*, Nature Photonics (2023), DOI: 10.1038/s41566-023-01212-1. 3. Tian Liman et al., *Research progress on measuring path integral propagators*, Acta Physica Sinica (2023), DOI: 10.7498/aps.72.20230902. 4. Zhu Shiliang team, *Direct Experimental Verification of the Feynman Path Integral via Photon Propagator Measurement*, Nature-family journal (August 2026); details pending final publication.

> Note: Figures such as the path count (1,419,857) and experimental details are cross-checked from NetEase reports, the university's announcements, and the 2023 paper; specifics should be confirmed against the final published 2026 paper.