Sample-half-inserted quantum interferometer

AI summary

Overview

Research area: Quantum photonics and quantum metrology — specifically Hong–Ou–Mandel (HOM) two-photon interference applied to measuring optical path differences in transparent samples.

Technical level: Intermediate (the conceptual picture of two-photon interference is accessible; the Fisher-information analysis and the interference model require some quantum optics background).

Scope (one sentence): The paper proposes, models, and experimentally demonstrates a "sample-half-inserted" HOM (SHOM) interferometer that raises the Fisher information per interference event by five orders of magnitude and measures optical path differences with nanometre (attosecond-scale) precision using only O(10^7) photons.

What This Paper Is About

Standard HOM interferometry can reach attosecond-scale timing resolution, but a single interference event carries very little Fisher information, so tens of thousands of repetitions and enormous photon numbers are needed to reach that precision (the authors cite Lyons et al.'s attosecond demonstration, which used 35,000 repetitions and O(10^11) photons). The goal of this work is to restructure the interferometer itself — inserting the sample into only half of the signal beam path rather than the whole path — so that a single measurement carries far more information about the sample's optical delay.

Key Contributions

1. A new interferometer geometry. The authors introduce the sample-half-inserted HOM (SHOM) interferometer, in which half of the signal photons pass through the sample and half travel through air. This asymmetric photon–sample interaction produces a distinctive dip–bump–dip coincidence pattern, derived analytically as a coincidence probability P_SHOM(τ, T) with an extra interference term absent from the standard HOM expression.

2. A five-orders-of-magnitude Fisher-information gain. Through Fisher-information and Cramér–Rao analysis, the SHOM configuration is shown to yield a Fisher information of approximately 5 × 10^6 ps^-2, versus roughly 30 ps^-2 for the conventional fully-inserted HOM arrangement, corresponding to about a two-order-of-magnitude improvement in achievable measurement precision.

3. Experimental validation with nanometre/attosecond metrics. Using a type-II SPDC source pumped at 405.0(1) nm, the authors measured optical path increments with an average precision of 4.09 nm (13.63 as) and an average accuracy of 1.22 nm (4.07 as) using O(10^7) photons.

4. Reframing a known artifact as a resource. The central bump in the dip–bump–dip structure is usually treated as an artifact in quantum optical coherence tomography (QOCT) and is often suppressed by pump-frequency dithering, broadband pumps, or machine learning. This work turns that feature into the metrological signal.

Main Findings

• Standard HOM baseline: Without a sample, the measured HOM dip has a full width at half maximum of 0.87(2) ps, centred at zero delay. With the glass sample fully inserted into the signal arm, the dip centre shifts from 0 to 3.868(7) ps.

• Distinctive SHOM pattern: In the half-inserted configuration the coincidence measurement shows two dips plus a central structure. As the sample temperature was raised from 28 °C to 52 °C, that central feature evolved from a bump into a dip.

• Temperature-to-optical-path calibration: Fixing the delay τ at the central interference position and scanning the sample temperature from 22 °C to 62 °C produced a sinusoidal coincidence pattern. Fitting gave a period of 24.03(2) °C, corresponding to an optical path increment of 405.0(1) nm; consequently a 2 °C increase corresponds to an optical-path increase of 33.7(3) nm, which the authors adopt as the "true" value.

• Measured Fisher information: The Fisher information calculated from the experimental interference data reaches a maximum of 3.40(6) × 10^6 ps^-2 — lower than the ideal theoretical 5 × 10^6 ps^-2 because of practical conditions such as imperfect visibility, but still five orders of magnitude above the standard HOM case.

• Estimated versus true optical path change: From 200 repeated trials (1 s per trial) at 38 °C and 40 °C, the estimated optical path change was c × ΔT = 33.89 ± 3.06 nm, compared with the known set value of 33.7(3) nm.

• Accuracy, precision, and stability: Repeating the procedure ten times gave an average accuracy of 1.22 nm (4.07 as); the reported average precision is 4.09 nm (13.63 as). Allan deviation averaged from 20 s to 70 s was 0.87 nm. The paper notes that the attosecond figure is inferred from delay measurements in the SHOM interferometer, not from direct photon timing.

• Head-to-head comparison (Table 1): Ref. [24] (HOM) used 35,000 trials and O(10^11) photons with average accuracy 1.7 nm and average precision 4.8 nm; this work (SHOM) used 200 trials and O(10^7) photons with average accuracy 1.22 nm and average precision 4.09 nm — 175 times fewer experimental trials.

• Prior benchmark context: The authors cite Branning et al.'s HOM measurement of birefringent crystal thickness with 0.1 fs (33 nm) resolution, and note that QOCT resolution has remained in the femtosecond regime.

• Scope caveat: The demonstration measures the sample's optical path difference, not its thickness; extracting thickness would require systematic characterization of the thermal expansion coefficient and the thermal-optic coefficient.

• Ultimate limit: The Cramér–Rao bound implies precision scales with the inverse square root of the number of repetitions; the limiting factors are instrumental — stepper-motor resolution and repeatability, temperature-control stability, and APD/TIA timing jitter — rather than the SHOM method itself.

Methodology in Plain English

The team built a two-photon interference experiment. A narrowband 405.0(1) nm laser pumps a PPKTP crystal, producing orthogonally polarized signal and idler photons by type-II spontaneous parametric down-conversion. The photons go through a delay system built from a polarizing beam splitter, two quarter-wave plates set at 45°, and two mirrors, with one mirror on a stepper motor to scan the optical delay τ. After the delay line, the photons pass a half-wave plate at 22.5° and a second polarizing beam splitter, are coupled into single-mode fibres, and are detected by avalanche photodiodes connected to a time-interval analyser, which records coincidence counts.

The key change is geometric: the glass sample is placed so that only half the signal beam passes through it while the other half goes through air. The sample sits on a Peltier-effect thermoelectric heater, so raising the temperature changes its optical path length in a known, roughly linear way — this provides a calibrated "true" value against which the interferometric estimate can be checked.

The analysis uses classical Fisher information computed from the coincidence probability as a function of the sample-induced delay, combined with the Cramér–Rao bound, to quantify how much information each measurement carries. Coincidence counts at fixed delay were recorded over 200 one-second trials at 38 °C and 40 °C, and the transmission-time difference was converted into an optical path change and compared with the known value.

Why This Matters

Impact on research: The work shows that the information content of a quantum interference measurement can be increased by engineering where the sample sits in the interferometer, not just by adding photons or repetitions. It provides an experimentally simple route to attosecond-scale delay sensitivity with a modest photon budget, and it reverses the usual treatment of the central QOCT "artifact," which has previously been something to suppress.

Real-world applications (as identified by the authors):

• Quantum optical coherence tomography (QOCT).
• Surface profiling for microscale manufacturing.
• Thickness monitoring of thin-film deposition in optical coating and semiconductor fabrication.
• Practical quantum-enhanced thickness measurement of transparent materials more generally.

Industry relevance: Nanometre-level, phase-insensitive optical path metrology with low photon counts is attractive for inline process control in thin-film deposition and semiconductor fabrication, and for non-contact inspection of transparent components, where stability against phase fluctuations and dispersion cancellation are practical advantages.

Future Directions

• Extending "sample-half-inserted" operation to other interferometers: The authors suggest adapting the strategy to N00N-state or Franson interferometers as a general metrology recipe.
• From optical path difference to true thickness: Converting the measured optical path into a thickness requires independent characterization of the sample's thermal expansion and thermal-optic coefficients — a systematic characterization step not performed here.
• Instrumentation-limited precision: Because the Cramér–Rao bound allows precision to improve with more repetitions (while noise stays white), better stepper motors, temperature control, and detector/analyser timing jitter are the concrete paths to better performance.
• Robustness and comparison: The paper discusses tolerance to variations in photon splitting ratio (SM Section XIII) and notes a related study by Lualdi et al. reporting attosecond-level precision with highly nondegenerate energy-entangled states, leaving open a direct comparison between the two approaches.

Target Audience

Quantum optics and quantum metrology researchers; photonics engineers working on precision measurement, OCT, or thin-film process control; and graduate students who already understand two-photon interference and want a worked example of Fisher-information analysis applied to a real interferometer. Readers without quantum optics background will find the conceptual idea accessible but the modelling sections demanding.

Authors’ abstract

Quantum technologies have been widely recognized as unprecedented opportunities for ultra-high precision metrology. As a celebrated example in modern quantum optics, the Hong-Ou-Mandel (HOM) interferometer is well-known for enabling temporal resolutions on the attosecond scale. However, the relatively low Fisher information per trial in ordinary HOM measurements typically necessitates tens of thousands of repetitions to achieve such precision. Here, we propose and demonstrate a sample-half-inserted HOM (SHOM) interferometer, which enhances the Fisher information by five orders of magnitude in a single interference event. By introducing an asymmetric photon-sample interaction, the SHOM configuration produces a distinctive dip-bump-dip interference structure, converting what was previously viewed as an artifact into a helpful metrological resource. Experimentally, we measured the optical path difference with an average precision of 4.09 nm (13.63 as) and an average accuracy of 1.22 nm (4.07 as) using $O(10^7)$ photons. Our results establish SHOM interferometry as an efficient phase-insensitive approach, not only paving the way toward practical quantum-enhanced thickness measurement for transparent materials, but also serving as an elegant strategy to improve the performance of various quantum devices.