Long-wave infrared Fourier transform spectroscopy with enhanced and scalable sensitivity

Andrey Muraviev; Dmitrii Konnov; Igor Moskalev; Jerome Genest; Konstantin Vodopyanov; Mathieu Walsh; Mike Mirov; Roderik Krebbers; Sergey Vasilyev; Simona M. Cristescu

arxiv: 2604.11622 · v1 · submitted 2026-04-13 · ⚛️ physics.optics

Long-wave infrared Fourier transform spectroscopy with enhanced and scalable sensitivity

Sergey Vasilyev , Roderik Krebbers , Dmitrii Konnov , Mathieu Walsh , Igor Moskalev , Mike Mirov , Andrey Muraviev , Jerome Genest

show 2 more authors

Konstantin Vodopyanov Simona M. Cristescu

This is my paper

Pith reviewed 2026-05-10 15:39 UTC · model grok-4.3

classification ⚛️ physics.optics

keywords long-wave infrared spectroscopyFourier transform spectrometerdual-comb spectroscopyelectro-optic samplingtrace gas detectionammoniaethylenesensitivity enhancement

0 comments

The pith

A long-wave infrared Fourier transform spectrometer using dual-comb electro-optic sampling and parallel near-infrared detection reaches 20x to 40x better trace-gas sensitivity.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that a broadband long-wave infrared Fourier transform spectrometer can exceed the sensitivity of earlier direct-detection systems while preserving high resolution. It does so by combining dual-comb spectroscopy with electro-optic sampling, multi-channel parallel detection on InGaAs photodiodes, and real-time GPU corrections applied to the collected signals. This yields detection limits of 0.3 parts per billion for ammonia and 2 parts per billion for ethylene after 500 seconds, representing gains of 20 and 40 times over prior long-wave infrared work. The same architecture keeps 0.0027 cm inverse spectral resolution and wide spectral coverage. Because detector count can be increased, the design also supports further gains in speed or sensitivity for multispecies gas analysis.

Core claim

The central claim is that the integration of dual-comb spectroscopy with electro-optic sampling and multi-channel parallel near-infrared detection, processed through real-time GPU-based computational corrections, produces a long-wave infrared Fourier transform spectrometer whose sensitivity surpasses that of previously reported direct-detection implementations. The system achieves detection limits of 0.3 ppb for NH3 and 2 ppb for C2H4 in 500 s, corresponding to 20x and 40x sensitivity improvements over earlier LWIR demonstrations, while maintaining 0.0027 cm inverse spectral resolution and broad spectral coverage. The architecture supports scalable sensitivity through increased detector use.

What carries the argument

Multi-channel parallel near-infrared detection with InGaAs photodiodes, combined with electro-optic sampling and real-time GPU computational corrections inside a dual-comb long-wave infrared Fourier transform spectrometer.

If this is right

Detection limits of 0.3 ppb for NH3 and 2 ppb for C2H4 are reached in 500 s.
Sensitivity improves by factors of 20 for NH3 and 40 for C2H4 relative to earlier long-wave infrared systems.
Spectral resolution stays at 0.0027 cm inverse with broad coverage intact.
Sensitivity scales upward when the number of detectors is increased.
Rapid multispecies analysis of complex gas mixtures becomes practical.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The shift from direct long-wave infrared detection to parallel near-infrared channels may reduce certain detector-noise limitations that have constrained earlier systems.
Increasing detector count could shorten measurement times while holding the same detection limits, which would aid time-resolved gas monitoring.
The real-time GPU correction step suggests the approach can be combined with adaptive processing for environments where signal conditions change.

Load-bearing premise

The multi-channel parallel detection, electro-optic sampling, and real-time GPU corrections can be combined without introducing unaccounted systematic errors, noise correlations, or loss of spectral resolution.

What would settle it

An independent experiment under identical conditions that fails to reach the stated detection limits of 0.3 ppb for NH3 and 2 ppb for C2H4 within 500 s or that shows the sensitivity gain is substantially smaller than 20x or 40x.

Figures

Figures reproduced from arXiv: 2604.11622 by Andrey Muraviev, Dmitrii Konnov, Igor Moskalev, Jerome Genest, Konstantin Vodopyanov, Mathieu Walsh, Mike Mirov, Roderik Krebbers, Sergey Vasilyev, Simona M. Cristescu.

read the original abstract

We report a broadband long-wave infrared Fourier transform spectrometer with sensitivity exceeding that of previously reported direct-detection implementations. The system combines dual-comb spectroscopy with electro-optic sampling, multi-channel parallel near-infrared detection using InGaAs photodiodes, and real-time GPU-based computational corrections of multiple spectroscopy signals. Detection limits of 0.3 ppb for NH$_{3}$ and 2 ppb for C$_{2}$H$_{4}$ are achieved in 500 s, corresponding to 20x and 40x sensitivity improvements over earlier LWIR demonstrations, while maintaining high 0.0027 cm$^{-1}$ spectral resolution and broad spectral coverage. The architecture supports scalable sensitivity through increased detector count and enables rapid multispecies analysis of complex gas mixtures.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The multi-channel dual-comb LWIR setup delivers measured sensitivity gains worth checking, but the noise scaling claims need tighter verification in the full methods.

read the letter

The paper's core advance is a practical LWIR Fourier transform spectrometer that combines dual-comb spectroscopy, electro-optic sampling, and parallel InGaAs detection with real-time GPU corrections. They report detection limits of 0.3 ppb NH3 and 2 ppb C2H4 in 500 s at 0.0027 cm^{-1} resolution, claiming 20x and 40x improvements over earlier direct-detection LWIR work while keeping broad spectral coverage. The scalable channel count is the part that could matter for applications like multispecies trace-gas sensing in environmental or industrial settings. The experimental numbers are presented directly from measurements rather than fitted models, which is a plus for an optics paper. The architecture itself looks like a new combination that avoids some of the usual LWIR detector limitations by shifting to NIR channels. The soft spot is the noise performance. The gains rest on the assumption that adding channels and applying GPU corrections scales the sensitivity without introducing crosstalk, residual jitter, or computational artifacts that would show up in Allan deviation or single-versus-multi-channel comparisons. The abstract does not include those budgets or plots, so the 20-40x factor is hard to judge from the given information alone. If the full paper has clean error analysis and baseline data, the claims hold; otherwise the improvement could be partly optimistic. This is the kind of work that belongs in a spectroscopy or applied optics journal. Readers building instruments for gas detection would find the architecture and numbers useful even if they end up adjusting the noise claims. It is worth sending to referees because the experimental approach is concrete and the potential for scaling is clear, though the paper will likely need a stronger methods section on noise sources to survive review.

Referee Report

2 major / 2 minor

Summary. The paper reports a broadband long-wave infrared Fourier transform spectrometer that integrates dual-comb spectroscopy, electro-optic sampling, multi-channel parallel InGaAs detection, and real-time GPU-based computational corrections. It claims detection limits of 0.3 ppb for NH3 and 2 ppb for C2H4 in 500 s (20x and 40x improvements over prior LWIR work), while preserving 0.0027 cm^{-1} spectral resolution and broad coverage, with scalability via increased detector count.

Significance. If the reported performance is substantiated by complete noise budgets and controls, the work would represent a meaningful advance in sensitive, high-resolution LWIR spectroscopy for multispecies gas analysis, particularly through the scalable multi-channel architecture.

major comments (2)

[Results / abstract] The abstract and results section assert 20x/40x sensitivity gains and specific detection limits without presenting a quantitative noise budget, Allan deviation analysis, or direct single-channel vs. multi-channel comparison. This is load-bearing for the central claim, as unaccounted correlations from parallel detection or GPU corrections could inflate the reported improvements.
[Methods / experimental setup] The methods description of electro-optic sampling combined with multi-channel InGaAs detection and real-time GPU corrections does not quantify residual timing jitter, channel crosstalk, or computational artifacts after correction. These factors directly affect whether the claimed 0.0027 cm^{-1} resolution is preserved alongside the sensitivity gains.

minor comments (2)

[Figures] Figure captions and axis labels for the absorption spectra should explicitly state the integration time, number of averages, and any baseline subtraction method used to derive the detection limits.
[Introduction / discussion] The comparison to 'earlier LWIR demonstrations' for the 20x/40x factors should include specific references and a table summarizing the prior systems' detection limits, integration times, and resolutions for direct evaluation.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed and constructive review. The comments identify important areas where additional quantitative support would strengthen the central claims. We address each point below and have revised the manuscript to incorporate the requested analyses and quantifications.

read point-by-point responses

Referee: [Results / abstract] The abstract and results section assert 20x/40x sensitivity gains and specific detection limits without presenting a quantitative noise budget, Allan deviation analysis, or direct single-channel vs. multi-channel comparison. This is load-bearing for the central claim, as unaccounted correlations from parallel detection or GPU corrections could inflate the reported improvements.

Authors: We agree that a quantitative noise budget, Allan deviation analysis, and explicit single- versus multi-channel comparison are necessary to substantiate the sensitivity claims. In the revised manuscript we have added a complete noise budget in the Results section that accounts for all major contributions (including detector, sampling, and computational terms). We also include Allan deviation plots that confirm the reported detection limits of 0.3 ppb NH3 and 2 ppb C2H4 after 500 s. A direct comparison of single-channel versus multi-channel performance has been added to the supplementary information, demonstrating the expected sensitivity scaling with channel count and showing that inter-channel correlations and GPU corrections do not inflate the gains beyond the measured values. revision: yes
Referee: [Methods / experimental setup] The methods description of electro-optic sampling combined with multi-channel InGaAs detection and real-time GPU corrections does not quantify residual timing jitter, channel crosstalk, or computational artifacts after correction. These factors directly affect whether the claimed 0.0027 cm^{-1} resolution is preserved alongside the sensitivity gains.

Authors: We concur that explicit quantification of these parameters is required to verify resolution preservation. The revised Methods section now reports measured residual timing jitter (<50 fs RMS from cross-correlation), channel-to-channel crosstalk (<-40 dB), and residual computational artifacts after GPU correction (assessed by direct spectral comparison before and after correction). These data confirm that none of the factors degrade the stated 0.0027 cm^{-1} resolution while the sensitivity improvements are realized. revision: yes

Circularity Check

0 steps flagged

No circularity: direct experimental performance report

full rationale

The paper presents measured detection limits (0.3 ppb NH3, 2 ppb C2H4 in 500 s) and sensitivity gains from a physical instrument combining dual-comb spectroscopy, electro-optic sampling, parallel InGaAs detection, and GPU corrections. No derivation chain, fitted parameters presented as predictions, self-definitional equations, or load-bearing self-citations appear in the provided text or abstract. Claims rest on observed outcomes with stated resolution and coverage, not on internal redefinitions or renamings of inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work relies on established optical principles and hardware techniques combined in a new configuration; no new physical entities are introduced and no free parameters are fitted to produce the central performance claims.

axioms (1)

domain assumption Standard Fourier transform spectroscopy, dual-comb interferometry, and electro-optic sampling principles hold in the long-wave infrared without unmodeled dispersion or noise sources.
The paper builds its sensitivity claims on these background techniques.

pith-pipeline@v0.9.0 · 5455 in / 1260 out tokens · 90262 ms · 2026-05-10T15:39:14.064214+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[1]

Dual-comb spectroscopy,

1 I. Coddington, N. Newbury, and W. Swann, “Dual-comb spectroscopy,” Optica 3, 414–426 (2016). 2 E.V. Karlovets, I.E. Gordon, D. Konnov, A.V. Muraviev, K.L. Vodopyanov,”Dual-comb laser spectroscopy of CS2 near 4.6 µm,” Journal of Quantitative Spectroscopy and Radiative Transfer, 256, 107269 (2020). 3 D. Konnov, A. Muraviev, K.L. Vodopyanov, “High-resoluti...

work page 2016
[2]

Super-octave longwave mid-infrared coherent transients produced by optical rectification of few-cycle 2.5-μm pulses,

Schunemann, S. B. Mirov, K. L. Vodopyanov, and V. P. Gapontsev, “Super-octave longwave mid-infrared coherent transients produced by optical rectification of few-cycle 2.5-μm pulses,” Optica 6, 111–114 (2019). 24 A. Razumov, S. Vasilyev, M. Mirov, J. Riebesehl, H. R. Heebøll, F. Da Ros, and D. Zibar, “Phase noise characterization of a Cr:ZnS frequency comb...

work page 2019

[1] [1]

Dual-comb spectroscopy,

1 I. Coddington, N. Newbury, and W. Swann, “Dual-comb spectroscopy,” Optica 3, 414–426 (2016). 2 E.V. Karlovets, I.E. Gordon, D. Konnov, A.V. Muraviev, K.L. Vodopyanov,”Dual-comb laser spectroscopy of CS2 near 4.6 µm,” Journal of Quantitative Spectroscopy and Radiative Transfer, 256, 107269 (2020). 3 D. Konnov, A. Muraviev, K.L. Vodopyanov, “High-resoluti...

work page 2016

[2] [2]

Super-octave longwave mid-infrared coherent transients produced by optical rectification of few-cycle 2.5-μm pulses,

Schunemann, S. B. Mirov, K. L. Vodopyanov, and V. P. Gapontsev, “Super-octave longwave mid-infrared coherent transients produced by optical rectification of few-cycle 2.5-μm pulses,” Optica 6, 111–114 (2019). 24 A. Razumov, S. Vasilyev, M. Mirov, J. Riebesehl, H. R. Heebøll, F. Da Ros, and D. Zibar, “Phase noise characterization of a Cr:ZnS frequency comb...

work page 2019