Enhanced quantum metrology with robust multipass interferometry

AI summary

Overview

Research area: Quantum metrology and quantum-enhanced sensing — specifically phase estimation with photonic interferometry under realistic photon loss.

Technical level: Intermediate to Advanced. The conceptual argument is accessible, but the quantitative analysis relies on quantum Fisher information (QFI), the quantum Cramér-Rao bound, and bosonic loss channels.

Scope (one sentence): The paper proposes and analyses a hybrid sensing strategy that embeds small, loss-tolerant entangled states inside a multipass Mach-Zehnder interferometer, showing via quantum and classical Fisher information that it beats conventional NOON-state and single-photon multipass schemes at fixed particle-pass budget under realistic loss.

What This Paper Is About

Quantum metrology normally buys precision by making entangled states larger, but large optical entangled states are extremely hard to produce and are fragile. Multipass interferometry offers a shortcut — one photon interacting repeatedly with the sample — yet it suffers the same exponential loss penalty, because the photon must survive every pass. The paper asks whether small, loss-resilient entangled states combined with repeated passes (a "hybrid" strategy) can retain robustness while still exploiting phase wind-up, and whether such a scheme is buildable with current photonic technology.

Key Contributions

1. A hybrid protocol. The authors combine small entangled probe states with a multipass cavity in one arm of a Mach-Zehnder interferometer, extending an existing lossy-state QFI formula to the multipass case by replacing the per-pass transmission η with η^p and multiplying the QFI by p².

2. A comparative loss analysis across state families. They evaluate NOON, dual-Fock ("BAT"), flat, a numerically optimised general state, and a new "rotated NOON" state, all at N = 4 and η = 0.95, ranking them by QFI per particle-pass, F_Q/(Np).

3. The rotated NOON state. They introduce a NOON state passed through a variable-reflectivity beam splitter whose reflectivity is optimised jointly with the number of passes, finding an optimum at p = 17 and reflectivity 0.83 for N = 4, η = 0.95.

4. A practical read-out analysis. They replace the abstract QFI with a concrete measurement — an adjustable output beam splitter followed by particle-number counting — and show the classical Fisher information per particle-pass preserves the ranking rotated NOON > BAT > NOON across the transmission values plotted.

Main Findings

• NOON and multipass are formally equivalent under loss. For NOON states the lossy multipass QFI is F_Q = 2N²p² η^{Np}/(1 + η^{Np}), where N and p appear only as the product Np — so particles and passes are interchangeable resources.

• All NOON sizes peak at the same value. In Fig. 2, F_Q/(Np) versus p for N = 1, 2, 5 at η = 0.95 reaches the same maximum in every case; smaller states simply need more passes, making them the cheaper route to the same performance.

• Loss-robust entangled states beat NOON. At N = 4 and η = 0.95, both the BAT state and the flat state achieve higher peak values of F_Q/(Np) than the NOON state (Fig. 3), though flat states have not yet been realised experimentally.

• Rotated NOON wins among buildable states. The rotated NOON state outperforms NOON and BAT and marginally outperforms the flat state (Fig. 4), and unlike the flat state it can be made with current technology.

• The advantage survives a realistic measurement. With an optimised output beam splitter plus number counting, Fig. 5 (η = 0.95) and Fig. 6 (maximised over p, plotted against η) show a clear, consistent hierarchy: rotated NOON dominates BAT, which dominates NOON, for all η displayed.

• The optimal pass number is finite. Because phase wind-up and cumulative loss compete, F_Q/(Np) rises to a maximum and then tails off, rather than improving indefinitely with more passes.

Methodology in Plain English

The authors work in the asymptotic regime of many independent measurements, so the quantum Cramér-Rao bound δφ ≥ 1/√(μ F_Q) sets the precision floor and everything reduces to computing the quantum Fisher information of the probe state.

The interferometer is a Mach-Zehnder with 50:50 beam splitters, two mirrors, and number-counting detectors, plus an optical cavity in the sensing arm so the probe passes the sample p times. Loss is modelled by treating the cavity mirrors as leaky with reflection probability η, i.e. a fictitious beam splitter that skims amplitude into an environment mode which is traced out; loss is assumed only in the sensing arm, since the reference arm can be engineered to be effectively lossless.

Starting from the known QFI expression for a general N-particle two-mode state under loss, they adapt it to multipass operation in two steps: the surviving fraction becomes η^p, and the coherently accumulated phase contributes a factor p². They then evaluate this expression for each candidate state, scaling by the number of particle-passes Np — the natural resource when the sample is delicate and flux must be minimised.

Finally, they step down from the QFI to the classical Fisher information obtainable from a real measurement: pass the state through a beam splitter of adjustable reflectivity and count particles at the outputs, optimising over the operating phase φ, the output reflectivity, and (for rotated NOON) the input reflectivity, separately for each p.

Why This Matters

The work shows a practical route to quantum-enhanced precision that does not require large entangled states — the central experimental bottleneck in optical quantum metrology — and it does so under realistic loss rather than the idealised lossless case. It reframes the relevant resource as particle-passes rather than particle number, which matters when the sample itself is the source of loss.

Real-world applications the paper situates this within (drawn from the introduction's survey of quantum-enhanced sensing):

• Gravitational-wave detection
• Atomic clocks and frequency standards
• Electric and magnetic field sensing
• Quantum imaging, and inertial sensing and navigation

Industry relevance: The proposal uses only linear optics, four-photon states that have already been demonstrated experimentally (both BAT and NOON), and standard photon counting, so it is aimed at near-term photonic sensing hardware rather than a future quantum-computing platform. Control over the pass number could reuse techniques from multipass microscopy.

Future Directions

• Larger entangled resources. The analysis deliberately restricts itself to N = 4 for practicality; the authors state further advantages follow once larger states can be generated.
• Better probe states. The numerically optimised general state of Eq. (8) is plotted as a benchmark but set aside as impractical to make — closing that gap is left open.
• Improved read-out strategies. The measurement scheme used is explicitly "non-optimised"; better practical POVMs approaching the QFI bound remain to be found.
• Experimental implementation. Building the cavity-based multipass element with controllable p, using multipass microscopy techniques, is the concrete next step the paper points to.

Target Audience

Quantum-metrology and quantum-optics researchers, particularly those working on photonic phase estimation, loss-tolerant probe states, and interferometer design; experimentalists in precision sensing and optical instrumentation; and graduate students with some grounding in quantum Fisher information who want a worked comparison of practical entangled probe states under realistic loss.

Authors’ abstract

Quantum metrology typically uses entangled states to achieve measurement precisions beyond the standard quantum limit. The advantage increases with the size of the entangled state, however generating and preserving large entangled states remains a major experimental challenge. Multipass protocols offer an alternative approach by allowing a single probe to interact repeatedly with the parameter of interest, but their performance is highly susceptible to loss, which accumulates over successive passes and rapidly erodes the quantum advantage. Here we introduce a hybrid strategy that combines small, loss-resilient entangled states with multipass interferometry. We show that this approach retains the robustness of small entangled probes while exploiting repeated interactions to achieve substantial enhancements in measurement precision. Furthermore, we propose a concrete implementation using currently available technologies, demonstrating that the predicted performance gains should be experimentally accessible with existing capabilities.