🔬 applied physics

Open-Path Methane Sensing via Backscattered Light in a Nonlinear Interferometer

This paper demonstrates robust open-path methane sensing at a distance of 4.6 meters under 60 dB loss by utilizing stimulated parametric down conversion in a nonlinear interferometer to detect diffusely backscattered mid-infrared light via near-infrared photons captured by a silicon CMOS camera, eliminating the need for highly reflective mirrors and precise alignment.

Original authors: Jinghan Dong, Weijie Nie, Arthur C. Cardoso, Haichen Zhou, Jingrui Zhang, John G. Rarity, Alex S. Clark

Published 2026-02-06

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jinghan Dong, Weijie Nie, Arthur C. Cardoso, Haichen Zhou, Jingrui Zhang, John G. Rarity, Alex S. Clark

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

Imagine you are trying to listen to a whisper in a noisy, windy room. Usually, to hear that whisper, you need a perfect, shiny mirror to bounce the sound back to you. But what if the object you are trying to listen to is a rough, matte wall, a crumpled piece of paper, or even a leaf? In the real world, most things don't act like mirrors; they scatter light in every direction, making it nearly impossible to catch a clear signal.

This paper describes a clever new way to "listen" for methane gas (a potent greenhouse gas) even when the light bounces off these messy, real-world surfaces. Here is how they did it, broken down into simple concepts:

The "Magic Trick" of Invisible Light

Usually, to detect methane, you need to shine a specific type of light (Mid-Infrared) through the air. Methane eats up this light, and by measuring how much is missing, you know how much gas is there. The problem? The detectors needed to see this Mid-Infrared light are expensive, need to be super cold, and are easily overwhelmed by background noise.

The researchers used a quantum physics "magic trick" called Stimulated Parametric Down-Conversion. Think of it like a translator:

They shine a "pump" laser into a special crystal.
They also shine a "seed" laser (the Mid-Infrared light that hunts for methane) into the same crystal.
The crystal acts like a factory, turning the seed light into a brand-new, "sibling" beam of light called a Signal.
Crucially, this new Signal beam is in the Near-Infrared range. This is the "cheap" range of light that standard, low-cost silicon cameras (like the ones in your phone) can see easily.

The Analogy: Imagine the Mid-Infrared light is a secret message written in a language only a few expensive experts can read. The crystal translates that message into English (Near-Infrared light). Now, a standard camera can read the message, even though the original secret language never touched the camera.

The Problem: The "Messy" Bounce

In previous experiments, scientists used perfect mirrors to bounce the Mid-Infrared light back to the crystal. But in the real world, you can't always put a mirror on a wall or a tree. If you shine light at a piece of white paper or a leaf, the light scatters everywhere (diffuse reflection). Most of it is lost, and the little bit that comes back is a mess, making it impossible to create a clear "Signal" beam to read.

The Solution: The "Catch-and-Return" Net

The team solved this by building a special two-lens system, acting like a sophisticated net.

The Setup: They aimed the light at a target 4.6 meters away (about 15 feet).
The Net: They used two lenses to catch the scattered light from the target and focus it perfectly back into the crystal.
The Match: They carefully calculated the size of the lenses so that the "messy" light coming back would fit perfectly inside the crystal, matching the size of the original beam.

The Analogy: Imagine trying to throw a ball back into a small cup from a distance. If you throw it randomly, it misses. But if you use a funnel (the lenses) to catch the ball and guide it straight into the cup, you can succeed even if the throw was messy. They tuned their "funnel" so that even the scattered light from a rough surface could get back into the crystal to do its job.

The Results: Seeing the Invisible

They tested this setup with different targets:

White Paper: A standard, rough surface.
A Leaf: A natural, uneven surface.
Brushed Metal and Glass: Surfaces that reflect light differently.

In all cases, the system worked. Even though the light lost a massive amount of energy (about 60 decibels, which is like whispering across a stadium) bouncing off these surfaces, the system could still detect the methane in the air.

They measured the "visibility" of the interference pattern (how clear the signal looked). Even as the distance increased and the signal got weaker, the system remained robust. They successfully measured background methane levels of about 3 parts per million, which is the natural level of methane in the air.

Why This Matters (According to the Paper)

The paper claims this is the first time this specific type of "undetected photon" sensing has been done using diffuse backscattering (bouncing off rough surfaces) instead of perfect mirrors.

No Mirrors Needed: You don't need to set up a perfect mirror on a wall or a tree to detect gas leaks. You can just point the sensor at the object.
Cheap Detectors: Because they translate the light to a range silicon cameras can see, they don't need expensive, cooled detectors.
Real-World Ready: It works on leaves, metal, glass, and paper, proving it can handle the messy reality of the outdoors.

In short, they built a system that can "see" gas leaks by looking at the light bouncing off ordinary objects, using a quantum trick to turn invisible, expensive-to-detect light into something a cheap camera can easily read.

Technical Summary: Open-Path Methane Sensing via Backscattered Light in a Nonlinear Interferometer

Problem Statement
Accurate methane detection is vital for monitoring emissions from natural gas infrastructure, agriculture, and anthropogenic sources to mitigate climate change and prevent safety hazards. While optical sensors offer advantages over semiconductor and electrochemical alternatives—such as real-time measurement, resistance to humidity, and low cross-sensitivity—most existing optical methane sensors rely on the 1.66 µm absorption band due to the cost and noise benefits of shortwave infrared (SWIR) detectors. However, the mid-infrared (MIR) region offers absorption features for methane that are over 60 times stronger than at 1.66 µm and are unaffected by common atmospheric components like water vapor.

Current nonlinear interferometry techniques based on spontaneous parametric down-conversion (SPDC) enable sensing with undetected photons but face significant limitations for outdoor, open-path applications. These systems typically require highly reflective mirrors and precise optical alignment to return idler photons. In realistic scenarios involving low-reflectivity or rough surfaces (e.g., walls, paper, natural surfaces), the idler photons suffer significant transmission losses. This degrades interference visibility and signal-to-noise ratio (SNR), making it difficult to detect visibility changes caused by gas absorption. Furthermore, SPDC-based systems are often limited by short coherence lengths and broadband spectral resolution.

Methodology
The authors propose a system based on Stimulated Parametric Down-Conversion (ST-PDC) to overcome these limitations. The experimental setup utilizes a Michelson-type nonlinear interferometer containing a magnesium oxide-doped periodically poled lithium niobate (MgO:PPLN) crystal.

Generation and Seeding: A 671 nm continuous-wave (CW) pump laser and a tunable 3.2 µm seeding laser are focused into the crystal. The seed stimulates the generation of a near-infrared (NIR) "Signal 1" (848 nm) and a MIR "Idler 1."
Open-Path Probing: The MIR Idler beam is directed toward a target. Instead of a mirror, the system utilizes diffuse backscattering from a Lambertian surface (e.g., white paper) or real-world materials. The backscattered light (Idler 2) is collected and returned to the crystal to stimulate a second PDC process, generating "Signal 2."
Detection: The interference between Signal 1 and Signal 2 is detected using a silicon-based CMOS camera at the NIR wavelength. This allows the use of low-cost, low-noise silicon detectors while probing methane in the MIR.
Optical Design: A critical component is a two-lens collection system designed to ensure spatial mode matching between Signal 1 and Signal 2. The authors simulate and optimize the focal lengths of the collection lenses to maintain the beam diameter and depth of focus of the backscattered Idler within the crystal, even as the target distance varies. This ensures that Signal 2 maintains the same spatial mode as Signal 1, preserving interference visibility despite high transmission losses.
Data Processing: By scanning the seeding laser wavelength across on- and off-resonance conditions for methane, the system measures changes in interference visibility. The transmittance is calculated from the visibility, and background methane concentrations are retrieved by fitting the absorption spectrum to the HITRAN database.

Key Contributions

First Demonstration of Undetected Backscattered Sensing: The work presents the first demonstration of background gas sensing using undetected, diffusely backscattered MIR light in a nonlinear interferometer.
Robustness to High Loss and Diffuse Targets: The system successfully detects methane at a distance of 4.6 meters (9.2 m round-trip path) under a total loss of 60 dB, using a Lambertian target. This contrasts with SPDC-based methods that typically fail under such high-loss, low-reflectivity conditions.
Mode-Matching Optimization: The authors identify and validate the critical role of lens selection in a two-lens collection system. They demonstrate that improper focal lengths lead to rapid signal degradation due to mode mismatch, whereas optimized lenses (specifically $f_1 = 18$ mm) allow the system to maintain interference visibility that scales as $1/L_0$ (where $L_0$ is distance), despite system loss scaling quadratically ( $1/L_0^2$ ).
Real-World Material Applicability: The method is extended beyond idealized white paper to include diverse real-world targets, including a leaf, glass, and a brushed metal plate, demonstrating robust sensing across various surface types.

Experimental Results

Detection Performance: The system achieved background methane detection at distances of 1.4 m, 3.1 m, and 4.6 m. The retrieved concentrations were consistent with ambient levels: $(2.96 \pm 0.29)$ ppm, $(3.10 \pm 0.32)$ ppm, and $(2.89 \pm 0.34)$ ppm, respectively.
Sensitivity: The minimum detectable methane concentration change is approximately 300 ppb, with a minimum detectable concentration of 8.3 ppm·m.
Angular Robustness: Comparative tests between a mirror and a Lambertian target showed that while mirror-based visibility drops to near zero with a $3^\circ$ rotation, the Lambertian target maintains detectable visibility up to $48^\circ$ . This confirms the system's suitability for applications where precise alignment is difficult.
Signal-to-Noise: At 4.6 m, the SNR of the visibility measurement was approximately 40, despite the high loss. The system successfully filtered out thermal background noise through the "single-mode" phase-matching of the ST-PDC process.

Significance and Claims
The paper claims to demonstrate a significant step forward in the practical application of nonlinear interferometry for outdoor sensing. By replacing the requirement for high-reflectivity mirrors with a diffuse backscattering scheme and optimizing for spatial mode matching, the authors enable methane detection in high-loss environments where previous SPDC-based methods would fail.

The authors modestly note that while they achieved detection at 4.6 m with a Lambertian target, the maximum distance is currently limited by the visibility dropping below detectable levels. They suggest that future improvements could involve balancing the interferometer by adjusting pump polarization or increasing seeding power to further enhance visibility. However, they caution that increasing seeding power requires careful management to avoid overwhelming the signal with spontaneous PDC background noise. The work establishes the feasibility of using undetected photons for open-path sensing on real-world surfaces, opening pathways for detecting other gases and potential imaging applications based on backscattered light.